Dell Technologies has launched a new storage platform in its PowerStore range that it claims offers "breakthrough" performance and efficiency.

Dell PowerStore Elite, as the new product is called, comes in three new models – Dell PowerStore 1500, Dell PowerStore 5500, and Dell PowerStore 9500. Depending on the model, Dell PowerStore Elite offers up to three times more performance compared to the PowerStore 1200T model and three times more throughput than the 9200T models, according to preliminary internal testing.

Varun Chhabra, SVP of infrastructure (ISG) and telecom marketing at Dell Technologies, said PowerStore Elite is the "most substantial update for the [PowerStore] platform in the last two years" and that it "represents the biggest leap forward in the platform's history".

"It's a full refresh of hardware and software, and builds on the success of PowerStore Prime from two years ago," he told press ahead of the announcement. "[It's] designed for what customers are facing next, and its timing is really, really important as... infrastructure planning is becoming far less predictable."

Dell PowerStore Elite specs and features

In terms of specs and features, Dell PowerStore Elite offers:

• Xeon Scalable processors, with up to 50% more CPU cores than PowerStore 3200T and PowerStore 5500
• DDR5 memory
• PCIe Gen 5 support
• 200Gb RDMA node interconnect for better load balancing and failover
• Up to 40 drives and 5.8PB effective capacity in a single 3U chassis
• Up to 3x density than prior generations on low-profile E3 NVMe flash
• Up to 40 network ports per appliance with 64Gb FC and 200/400Gb Ethernet-ready connectivity.

It also offers a guaranteed 6:1 data compression.

AI-based efficiency improvements

Continuing the theme of Dell Technologies World, AI is a key feature in PowerStore Elite. According to the company, built-in AI continuously balances workloads, tunes performance, and improves efficiency, which, when put together, reduces manual effort by 95% compared to traditional array management. It also offers Dell Private Cloud integration.

"PowerStore Elite is not just a faster array, it is a brand new class of a modern data platform that is built to help customers lead through change, not just react to it," Chhabra said.

PowerStore Elite will be available worldwide from July 2026.

Dell Technologies has unveiled a new, integrated rack-scale product for AI workloads that it claims can go from delivery to running live AI or HPC workloads in as little as six-and-a-half hours.

Dell PowerRack was announced at the company’s annual Dell Technologies World conference in Las Vegas as one of several updates to Dell AI Factory with Nvidia that were announced on the opening day keynote.

“One of the most common challenges customers have with AI infrastructure is dealing with the complexity of bringing all the pieces together,” Varun Chhabra, SVP of infrastructure (ISG) and telecom marketing at Dell, told journalists and analysts ahead of the conference. Dell PowerRack, he said, brings together AI compute, AI networking, and AI storage as scalable units designed, tested, and validated as one system.

“With Dell PowerRack, customers no longer need to buy components and hope they work together. They can deploy a fully integrated system in which the thermal design, the power management, and the software for infrastructure management have all been engineered to work as one unit,” he said.

Dell PowerRack is available in compute, networking, and storage configurations.

PowerRack for Networking has over 800 Tb/sec of switching capacity with eight Dell PowerSwitch SN6600 LG Ethernet switches per rack, which Chhabra said enables “the high-volume, low-latency east-west traffic required by GPU-dense environments”.

PowerRack for Storage, meanwhile, is a software-defined system based on Dell Exascale Storage and incorporates Dell PowerFlex. The addition of PowerFlex brings block storage into the equation. This means it can be used not only for AI and HPC workloads but also for very demanding but more traditional enterprise workloads.

Dell PowerRack for storage is, according to Chhabra, “the only 4-in-1 storage that's built for extreme-scale HPC, HPC AI, and demanding enterprise workloads”.

ITPro’s managing editor Jane McCallion is at Dell Technologies World all week. Click here to follow her live blog, or here for all our coverage of the event, including news and analysis.

Chinese manufacturer LincPlus started life producing budget-priced and aesthetically pleasing laptops and tablets, and moved its focus onto the NAS market a couple of years ago. Its first product in 2024 was the LincStation N1, which boldly claimed to be the world's first six-bay all SSD appliance, followed by the N2 the following year, which employed the same compact chassis and delivered a big performance boost.

In this review, we look at the LincStation E1 – a table-top NAS which teams up two LFF/SFF SATA hot-plug drive bays with a pair of M.2 NVMe SSD slots. Recently launched on the Kickstarter crowdfunding platform, its biggest claim to fame is its very low price, with the first 200 units going at a super-early-bird price of €129, followed by an early-bird price of €149 and then a regular price of €219, which pans out to around £190.

Perhaps the biggest change is in software; unlike the N1 and N2, which both employ Lime Technology's Unraid software, the E1 introduces LincPlus' very own LincOS operating system. Unraid is undoubtedly a powerful OS with plenty of features, but it isn't the easiest to configure, and LincOS aims to deliver a more user-friendly experience.

LincPlus LincStation E1 review: Design and hardware features

The E1 employs a solidly constructed plastic chassis with the two telescopic tool-free hot-swap drive carriers accessed by removing the magnetic upper cover. In the base, you'll find a non-magnetic cover with the two M.2 NVMe slots behind it and a single fan sitting directly on top of them.

It's all change in the CPU department as LincPlus moves away from Intel, with the E1 receiving a quad-core 2GHz Rockchip RK3568 SoC sporting an ARM Mali G52 GPU and a 1TOPs NPU (neural processing unit). Memory is the DDR4 variety with the 4GB soldered on the motherboard and not upgradeable.

There was some confusion when the appliance was first launched, as its KickStarter project page showed the network port as Gigabit, and LincPlus initially confirmed that the 2.5GbE port in our early pre-launch model had been dropped down to Gigabit on production models. We were subsequently advised that this was incorrect and that production appliances will have a 2.5GbE port.

There are plenty more hardware features, along with integrated dual-band Wi-Fi support, you have a USB-A 3.2 Gen 2 USB port at the front and two more USB-A 2.0 ports at the rear. The HDMI 2.1 port alongside receives 4K 60Hz video from the SoC's GPU.

LincPlus LincStation E1 review: Swift deployment

For testing, we fitted two 4TB Seagate IronWolf NAS HDDs, added a couple of Kingston M.2 NVMe SSDs, and made sure the NVMe bay cover had been pushed forward until it was locked in place. The appliance was then hooked up to the lab's Zyxel XMG1930-30HP multi-Gig PoE++ switch, which confirmed a 2.5GbE connection speed.

Multiple deployment options are available, as you can attach an HD monitor, keyboard, and mouse to the appliance and use the desktop user interface. The easiest method is to install the LincStation Windows, iOS, or Android client apps, which run a wizard for registering the appliance and setting up an admin account.

To create storage spaces, you select the Profile/My Device/Storage Management option from the client and choose your drives with RAID0, 1, 10, 5, and 6 arrays supported. We created separate RAID1 mirrored storage spaces for the HDDs and NVMe SSDs. The appliance doesn't support SSD caching in the traditional sense, but you can use NVMe SSDs as a fast repository for file sharing and downloads.

LincPlus LincStation E1 review: LincOS features

LincOS is easy to use, and we found all features can be accessed directly from the client apps, which open with quick access links to personal storage spaces and a collection of apps for file sharing, multimedia services, remote downloads, and more. The client Profile menu offers a desktop mode, which presents the same graphical interface as a locally attached monitor.

The admin user can load a basic web UI from the client with access to personal spaces for viewing, creating, and sharing folders, monitoring upload tasks, loading a recycle bin, creating encrypted spaces, and managing share spaces. New users are created from the client, and they can log in from their choice of client and use the web UI for accessing their personal space, task monitoring, and a recycle bin.

The Backup Disk tool secures PC and mobile data to the appliance, where you choose local folders, an appliance storage space, the file types to be backed up, and whether to use WiFi only or any network. The Sync Disk tool keeps selected folders on the local device and NAS in step, and in both cases, the processes are automated with an option to run jobs on demand as well.

The Secure Space feature allows users to create hardware-encrypted, password-protected areas in storage spaces for private data. Folders in this space cannot be shared, and the client provides an option for admins to back up encryption passwords for all user accounts for disaster recovery purposes.

Multimedia services are extensive as LincOS has music and video centre apps, a smart album for photos, and Photo Frame for creating slideshows. The smart album's local AI-powered photo image and facial recognition work well, while the smart library AI document centre provides full text search facilities for all file content.

LincPlus LincStation E1 review: Gigabit performance

With a share on the appliance's HDD RAID1 storage space mapped to a Dell PowerEdge Windows Server 2025 host on our 10GbE LAN, Iometer recorded uninspiring sequential read and write rates of 216MB/sec and 131MB/sec with the Windows client app showing CPU utilisation of up to 80%.

Real world numbers also failed to impress with 25GB file copies between the NAS and server delivering average read and write rates 170MB/sec and 112MB/sec while our 22.4GB backup test folder with 10,500 small files was secured using drag and drop at 66MB/sec.

Moving to a share on the RAID1 NVMe SSD mirror saw minor performance improvements with Iometer recording sequential read and write rates of 225MB/sec and 139MB/sec. Read and write speeds for our large file copies stepped up slightly to 174MB/sec and 117MB/sec while our backup test mustered 71MB/sec.

LincPlus LincStation E1 review: Is it worth it?

The KickStarter campaign ends at the start of June, and with a projected regular price of around £190, the LincStation E1 is probably the lowest-cost 4-bay NAS on the market. Performance over 2.5GbE is below expectations, but it should satisfy the target market of home and small business users.

A standout feature is the new LincOS software, as even though it's still in development, it already offers a good range of storage features with the smart LincStation clients making light work of configuration and management. LincPlus advised us that the next LincOS update will add an improved web GUI, a new task manager, drag and drop support for the cloud drive module, a redesigned music module, and the LincClaw AI assistant.

LincPlus LincStation E1 specifications

HPE has announced an expansion to its ProLiant edge portfolio, with a particular focus placed on supporting organizations operating in “harsh environments”.

The new HPE ProLiant Compute EL2000 chassis, which supports up to two HPE ProLiant Compute EL220 Gen12 servers or one EL240 Gen12 server, is designed specifically for applications in manufacturing, retail, telecommunications, and national security.

According to HPE, the new servers boast a range of performance improvements compared to prior generations, and can scale from eight up to 144 Intel Xeon 6 cores.

The EL240 Gen12 server is also available with Nvidia RTX Pro 4500 or Nvidia RTX Pro 6000 Blackwell server edition GPUs.

Durability is a key focus here, according to HPE. The new server range can operate in extreme temperatures ranging from -40 degrees Celsius to 55 degrees Celsius, as well as environments with up to 95% humidity.

They’re also designed to compensate for heavy vibration from aircraft or ground vehicles, as well as electromagnetic interference (EMI) and environmental contaminants.

During a press briefing ahead of the announcement, Krista Satterthwaite, senior vice president and GM of HPE’s compute group, said the focus on durability and ruggedness comes in direct response to changing edge requirements.

“Every edge is different, and edge is hard,” she said. “Servers at the edge need to be tough across industries like retail, manufacturing, [and] telecommunications.”

“Workloads are running at edge locations that are understaffed, constrained by space power, connectivity, and security becomes even more critical because these environments are much more exposed, which means protection has to be built in from the start.”

HPE touts telecoms, defense potential

HPE revealed that each platform is now available with an “Environmental Ruggedization” kit, enabling organizations operating in harsh conditions to compensate for “extreme stressors”.

“We're introducing an environmental ruggedized kit that enables these servers to handle extreme stressors,” Satterthwaite told assembled media.

“These new offerings satisfy customer needs for purpose built systems that deliver highly reliable performance in harsh and highly distributed locations that are designed to operate remotely with end to end visibility and control.

According to HPE, the ruggedization features adhere to a range of standards, including US national security standards, which require validated server survivability capabilities in extreme temperatures or high or low altitudes.

Satterthwaite noted that growing demands for edge capabilities in the defense industry and the public sector have played a key role in the durability focus.

“That is one of the main focus areas, which is why we have so many things that we’ve pursued in terms of certifications for more ruggedized environments,” she commented.

“So big, big demand coming from the defense space, in particular in the public sector.”

John Carter, VP for mainstream compute at HPE, echoed Satterthwaite’s comments, noting that the systems will “go anywhere” and “withstand any environment”.

“We’ve seen demand and had customer conversations from everything from space stations to battleships, to Jeeps, to anything else you can think of,” he said.

Elsewhere, the new range also adheres to telecoms network equipment standards that support 5G and radio access network (RAN) availability in unattended environments.

On the telecommunications front, Carter told assembled media this is an industry where the company is seeing surging demand for edge capabilities.

“Telcos have really been embracing the edge for quite a while,” he said. “They’ve been expanding the 5G roll-outs, now they’re starting to look at 6G in the future, and that is requiring a new level of performance and compute.”

“But we have seen numerous different environments from different customers. It just depends on the provider. And we do work with those telco providers from everything from the extreme edge all the way back to the packet core,” Carter added.

FOLLOW US ON SOCIAL MEDIA

Ugreen is rapidly gaining a strong reputation in the NAS market with its ever-expanding range of desktop appliances. And the latest DXP4800 Pro focuses on demanding home and small office users who want plenty of power and performance for their money. Stepping up as Ugreen's flagship 4-bay appliance, it takes everything we liked about the two-year-old DXP4800 Plus and gives it a performance boost.

The Gen12 5-core Intel Pentium Gold 8505 CPU in the Plus model gets replaced with a more modern Gen13 6-core Core i3-1315U offering four efficient plus two multi-thread performance cores and a top Turbo frequency of 4.5GHz. Memory stays the same at DDR5, but those with deep pockets can upgrade the base 8GB to 96GB – the maximum supported by this CPU.

Along with four front drive bays, you have a pair of M.2 NVMe SSD slots lurking in the base, and it's worth noting these are the faster PCIe Gen4 variety. Network port choices remain the same, although these are superior to a lot of the competition as the appliance sports 2.5GbE and 10GbE multi-Gigabit ports.

Ugreen NASync DXP4800 Pro review: Build quality and hardware features

No cheap plastic chassis here as the DXP4800 Pro is clothed in a sleek and sturdy aluminium alloy shell. SATA hard disks slip easily into the clever telescopic tool-free carriers, which also have mounting holes for SFF SATA SSDs.

The two M.2 slots are easy to access as the cover is spring-loaded, and you'll find two SO-DIMM slots next to them with the base 8GB of memory provided on one stick. General cooling is handled efficiently by a big rear 14 cm diameter fan behind a magnetic dust cover, and we found the appliance to be pleasingly quiet.

The appliance offers plenty of other interfaces with 10Gbps USB-C 3.2 and USB-A 3.2 at the front next to an SD Card reader slot. A peek around the back reveals one 5Gbps USB-A 3.2 and two USB-A 2.0 ports, while 4K 60Hz video from the CPU's embedded Intel UHD Graphics GPU is piped through to an HDMI 2.0 port.

We don't often comment on packaging, but Ugreen gets praise for all the extra bits and prices it includes in the box. The external PSU comes with a UK plug lead, and the accessory box contains two Ethernet cables, a screwdriver, and a full set of screws for SFF drives, two carrier locking keys, two adhesive silicon heat pads for M.2 devices, and even a Ugreen-branded HDMI cable.

Ugreen NASync DXP4800 Pro review: Deployment and UGOS apps

For testing, we fitted four 4TB Seagate IronWolf NAS HDDs, which were accepted without any problems, as unlike Synology, Ugreen doesn't have any drive restriction policies. Installation is a breeze as Ugreen's discovery web portal found the appliance on our network, initialised it, and downloaded the latest UGOS software.

The UGOS web console is nicely designed and aids storage pool creation by offering sage advice on the various RAID options. We chose the recommended RAID5 volume and were prompted to create personal and shared folders and configure remote access, which creates a unique Ugreenlink ID and secure web link.

During volume creation, you can choose from the Ext4 and Btrfs file systems, with the latter the preferred choice if you want snapshots, storage space quotas, and higher levels of data integrity protection. We also noted that the Samba file sharing service is disabled by default, which we think is a good security measure.

UGOS is light on apps as it only offers 29 for the DXP4800 Pro – well short of the huge range offered by Qnap and Synology. Data protection features start with the Snapshot app, which supports manual and scheduled snapshots on user and shared network folders stored in Btrfs volumes, but Ugreen has yet to update this to support iSCSI LUNs.

The Sync & Backup app secures local data to a remote Ugreen NAS and vice versa, and also supports Rsync-compliant servers. The app has a download link for Ugreen's NAS Windows and macOS desktop tool for creating sync tasks that secure selected data to a shared folder and update it in real time, while the Cloud Drives app provides file syncing for seven cloud storage services, including OneDrive and Google Drive.

The Security Manager app provides real-time and scheduled anti-malware protection with quarantining. Virtualization is on the menu with the Docker app running lightweight containers and the Virtual Machine app providing full hypervisor services.

Ugreen NASync DXP4800 Pro review: 10GbE performance

For performance testing, we hooked the NAS up to the lab's 10GbE network and mapped a share to a Dell PowerEdge Windows Server 2025 host. UGOS now supports 9000-byte Jumbo frames, and with these enabled, we saw good NAS speeds with Iometer reporting sequential read and write rates of 9.2Gbits/sec and 8.1Gbits/sec

For our real-world tests, copies of a large 25GB test file between the appliance and server delivered sustained read and write rates of 3.2Gbit/sec and 2.7Gbits/sec while backing up a 22.4GB folder with 10,500 small files averaged a reasonable 1.5Gbits/sec. Ugreen added the SAN Manager app a while ago, and moving to IP SANs saw a 500GB iSCSI target deliver Iometer read and write rates of 9.2Gbits/sec and 7.5Gbits/sec.

For cache testing, we fitted two 480GB Kingston M.2 NVMe SSDs and assigned them to our data volume as a mirrored read/write cache. This made noticeable improvements to random write operations with our NAS share and iSCSI LUN, seeing respective speed increases of 123% and 120%.

Ugreen NASync DXP4800 Pro review: Is it worth it?

With Ugreen currently offering the NASync DXP4800 Pro for £620 including VAT, it compares very well on value with the competition. Synology's DS925+, for example, costs a shade under £600 but only supports DDR4 memory, has dual 2.5GbE ports, is powered by a pensionable AMD Ryzen 1500B CPU, and doesn't offer a 10GbE upgrade option.

Ugreen's build quality is also vastly superior, but there's a lot of work to be done with the UGOS software before it can match the incredible range of apps, services, security, and data protection features offered by Synology's DSM and Qnap's QuTS Hero. The DXP4800 does deliver an impressive hardware package for the price, though, and if you can cope with the basic range of apps, is a great choice for home and office users seeking a powerful NAS appliance with great 10GbE performance.

Ugreen NASync DXP4800 Pro specifications

Fast-growing businesses seeking a tower server that can keep up with them will find HPE's ProLiant Compute ML350 Gen12 a worthy partner. Stepping up as HPE's most powerful tower yet, it supports dual Xeon 6 P-Core CPUs up to 86 cores plus a massive 8TB of DDR5 memory and presents a wealth of storage permutations.

The ML350 Gen12 looks capable of turning its hand to a multitude of tasks as its modular chassis offers an expansion potential that defies belief. Along with SMBs, HPE is targeting this server at a range of other scenarios, including ROBOs, edge deployments, and retail outlets, along with enterprise workloads such as IT infrastructure, data management, VDI, and ERP/CRM.

There's more, as it can handle up to eight single-width or four double-width Nvidia GPUs. These attributes make this server a fine choice for demanding AI, machine learning (ML), and database analytics tasks.

HPE ProLiant Compute ML350 Gen12 review: Design and expansion

Make no mistake, this server is big with the chassis measuring 46.2cms high, 71.2cms deep and 17.4cms wide, which equates to an internal volume of just over 57 litres. Build quality is impressively sturdy, with even the base model tipping the scales at a hefty 24kgs. HPE also offers a conversion kit that transforms it into a 4U rack server, which occupies 5U of rack height.

Slipping the lockable side panel off shows HPE had made excellent use of the server's voluminous interior. Behind the front storage bays is a full-height cooling fan array with base systems coming with three NHP (non-hot-pluggable) modules included.

The assembly has room for up to eight fans and the dual 2.5GHz Xeon 6730P CPUs in our system, and their 250W TDPs dictate that we need a full house of hot-plug modules. CPUs over 225W also require the high-performance heatsink kits, which extend large radiators into the fan array for maximum cooling and work extremely well, as our system was whisper-quiet.

Expansion potential is impressive as further back are two riser cages, each with up to four full-height, full-length PCIe Gen5 x16 slots. The quad-slot riser has on-board connectors providing backup power for HPE's storage controllers, while the dual-slot risers also have GPU sideband and auxiliary power connectors.

HPE's tool-free design comes into play as the two main riser cages can be easily released with just a flick of the wrist. The motherboard has two edge connectors for PCIe Gen5 OCP cards, and the standard PCIe slot count can be pushed to ten, as there's even space for a tertiary dual-slot riser in front of the upper PSU bays.

HPE ProLiant Compute ML350 Gen12 review: Storage features

As you'd expect, there's plenty of room at the front for storage, and the smart triple box design makes the server very flexible. Depending on your choice of backplane, each box supports four LFF SAS/SATA, eight SFF SAS/SATA/NVMe, or twelve EDSFF E3.S NVMe storage devices, and you can mix and match them as required.

The motherboard has Intel's Virtual RAID on CPU (VROC) for HPE ProLiant Servers Gen12 embedded, which is specifically for NVMe SSDs with direct to CPU connections and supports software-managed stripes, mirrors, and RAID5 arrays. For all other storage devices, there are HPE's Gen11 hardware RAID controllers with a choice of three OCP and three PCIe versions – we have the MR408i-o OCP card, which supports all the usual RAID array suspects and comes with 4GB of Flash-backed cache.

We also have HPE's NS204i-u V2 boot optimised storage device accessible at the rear and fitted with dual 480GB mirrored M.2 NVMe SSDs. If you want it hidden away from wandering fingers, you can have it front-mounted so it's behind the lockable bezel.

HPE ProLiant Compute ML350 Gen12 review: Remote management

The server comes with HPE's excellent iLO7 remote management controller, although, unlike most of the Gen12 rack servers, this is not implemented as an OCP-compliant DC-SCM (data center secure control module) board but embedded on the motherboard. No matter, as it provides all the same top-notch security features.

Platform security and supply chain integrity are assured, and the iLO7 silicon root of trust (RoT) digital firmware fingerprinting is future-proofed against next-generation quantum computing threats. It also includes HPE's patented Secure Enclave – a physically tamper-resistant security processor for storing sensitive data such as encryption keys, passwords, and security configurations.

The iLO7 web interface is designed around workflows, with the customisable home page dashboard presenting up to six cards for views of key areas such as host health and security issues. Colour-coded icons are provided on each card for at-a-glance status views, and you drill down into each one for more detailed information.

We run HPE's OneView in the lab on a Hyper-V host and had no problems importing the ML350 Gen12. The free OneView Standard licence supports server inventory and monitoring, with the Advanced licence offering OS deployment tools plus server, firmware, and power management services.

Enterprises will love HPE's GreenLake Compute Ops Management (COM) cloud service as it connects directly to the iLO7 controller and, unlike Dell's AIOps, doesn't require an intermediary app running on the local network. COM presents a complete overview of your server estate and provides tools for server deployment and management, automated system lifecycle compliance, and AI-powered insights.

HPE ProLiant Compute ML350 Gen12 review: Is it worth it?

Yes, very much so. We can see from HPE's One Config Simple portal that prices for base systems start at around £6,800, which gets you a 12-core Xeon 6505P CPU, 32GB of DDR5 Smart Memory, a four LFF drive box with two 8TB SATA drives, an MR408i-o OCP RAID card, and dual 800W PSUs.

The ProLiant Compute ML350 Gen12 offers a lot of tower server for your money and delivers an expansion potential that makes it incredibly versatile. Clearly capable of handling a diverse range of workloads, it's an ideal choice for businesses that want a high expansion and plenty of growth potential.

HPE ProLiant Compute ML350 Gen12 specifications

Lenovo has been a key player in the edge computing market for many years, and its latest ThinkEdge server family offers a range of innovative solutions. The ThinkEdge SE100 on review is a prime example, as it delivers entry-level AI inferencing to the network edge in a chassis that Lenovo claims is 85% smaller than a standard 1U rack server.

The SE100 is, indeed, one of the smallest edge servers we've yet seen as it measures a mere 53mm high with the fan shroud, 142mm wide, and 278mm deep. This allows it to fit into spaces few other servers can, and Lenovo's mounting options include desktop, VESA, and DIN brackets along with rack mount kits for fitting two or three of them alongside each other.

It's geared up for harsh environments as the chassis is shock and vibration resistant and has an IP50 dust protection rating. The SE100 is cleared for continuous operations in ambient temperatures of between 5 and 45 degrees C, and the optional PCIe expansion kits offer a rear MERV5-rated dust filter.

Targeting applications such as ML (machine learning), augmented reality, smart surveillance with object detection, and edge AI, the SE100 has plenty of power on tap as it's available with either an Intel Core Ultra 7 255H or Ultra 5 225H CPU, with both offering an embedded Intel Arc GPU and Intel AI Boost NPU (neural processing unit). Memory is the TruDDR5 variety with up to 64GB supported, and you can specify standard 5,600MHz SO-DIMMs or faster 6,400MHz CSO-DIMMs.

Lenovo ThinkEdge SE100 review: Build quality and features

The SE100 is built like a tank with the base node tipping the scales at a hefty 2.4kgs. The thick aluminium chassis acts as a heatsink with a ribbed underside and an upper finned shroud equipped with dual cooling fans.

The air shroud is designed to maintain optimum cooling even if something is placed on top of the server, and it also allows Lenovo to dispense with the micro-fans typically found in edge servers and replace them with larger 6.5cms diameter fans to reduce noise. They certainly work well, as during testing, we found the SE100 to be whisper-quiet.

A range of power options are available with our system, supplied with a single 140W external PSU which connects to a dedicated USB-C port at the rear. There's a second USB-C port alongside for adding a second, redundant PSU, and it also doubles up as a local management port for connecting a mobile device running the XClarity Controller (XCC) app.

The SE100 has two on-board video controllers with the embedded XCC's AST2600 chip piping 1080p output to the front USB-C port, which has integral DisplayPort support. The two HDMI 2.0 ports handle output from the CPU's Arc GPU and deliver 4K video at 60Hz.

Network connections are provided by dual embedded Gigabit ports, and there's a dedicated Gigabit port for XCC remote management access. You also have pairs of USB-A 3.2 Gen2 ports fore and aft and two USB-C 3.2 Gen2 at the rear with both supporting DisplayPort.

Lenovo ThinkEdge SE100 review: Storage and expansion

The SE100 offers a fairly basic storage proposition as the motherboard has three PCIe Gen4 M.2 slots, with the first one supporting 2280 SATA or NVMe SSD boot devices. The other two are for general data storage and have room for 2280 and 22110 M.2 NVMe SSDs, but note that RAID is not available for this server.

The base node has a PCIe Gen4 expansion connector on its side, and Lenovo offers two expansion kit options, which are attached to the side of the chassis and increase its width to 214mm. The GPU kit supports a single-width, actively cooled card, and you can choose from Nvidia Quadro RTX 2000E Ada or RTX A1000 versions.

The network expansion kit is available with Broadcom quad-Gigabit or dual-10GbE adapters and includes two internal fan modules. The 1U2N rack mount kit supports two base nodes with expansion kits, and this doesn't require the node fan shrouds which drop their height to 1U. The 1U3N rack kit allows three SE100 nodes without the expansion kits to be placed in a row.

Lenovo ThinkEdge SE350 V2 review: Server management

The SE100 has the same XClarity Controller (XCC) 2 chip and dedicated Gigabit port you'll find on Lenovo's standard ThinkSystem rack and tower servers. The base system comes with an XCC Standard licence, which enables features such as system health and power monitoring, while a Platinum upgrade costs around £220 and adds a host of extras, including remote KVM, virtual media services, remote OS deployment, power-capping, and the Neighbour Group feature for managing multiple servers from one XCC interface.

The XCC's web console presents plenty of information about critical hardware components along with a hardware inventory and firmware upgrade tools. It presents a power usage table for the system, CPU, memory, and other components, but temperature data is not available, and the main utilization page only shows the CMOS battery voltage and fan speeds.

The SE100 is currently certified for Windows 11 Enterprise, Windows 11 IoT Enterprise, RHEL, Ubuntu, and SUSE Linux Micro. We found OS deployment is a breeze as we used the XCC provisioning manager's Effortless Install option to load Windows 11 Enterprise in 30 minutes.

The chassis doesn't have the motion detection sensors found in larger ThinkEdge systems such as the SE350 V2, but tamper detection events can be linked to the XCC, and if a cover is removed, it will initiate a system lockdown mode. More platform security is evident as the SE100 employs Silicon Root of Trust for supply chain integrity, it supports SEDs (self-encrypting SSDs), and the XCC System Guard feature takes reference snapshots of the current system inventory and can stop the server from booting if changes are detected.

Lenovo ThinkEdge SE100 review: Is it worth it?

With prices for base nodes starting in the £3-4K price range, the ThinkEdge SE100 offers a tempting solution to SMBs that thought edge computing was beyond their budget. The Intel Core Ultra hybrid CPUs are quite capable of handling demanding tasks such as entry-level AI inferencing and the system can be easily expanded with a more powerful Nvidia GPU if needed.

Its rugged build and compact dimensions allow it to be deployed in environments that would be no-go areas for standard servers and the smart cooling design makes for silent running. The ThinkEdge SE100 is an impressive little edge server that will appeal to a wide range of businesses and the multiple deployment options makes it incredibly versatile.

Lenovo ThinkEdge SE100 specifications

Qnap offers an impressive range of enterprise all-Flash NAS appliances, and its TS-h1090FU turned heads when it was launched, as it delivers an impressive and very powerful package in a compact 1U rack chassis. A near-£9K starter price tag won't endear it to SMBs, though, and we review the new TS-h1077AFU, which aims to offer budget-conscious businesses a more affordable all-Flash alternative.

It employs the same short-depth chassis equipped with ten SFF front hot-swap bays, but similarities are only skin deep, as there are a lot of internal changes designed to make the price more palatable. The TS-h1077AFU only supports SATA SFF SSDs and not U.2 NVMe SSDs, and the dual embedded 25GbE ports in the TS-h1090FU have been replaced with 10GbE ports.

It's all change in the processing department as the TS-1077AFU receives an 8-core 3.8GHz AMD Ryzen 7 7700 CPU with embedded AMD Radeon Graphics. Another advantage of the Ryzen CPU is it supports fast DDR5 memory with the price including 32GB upgradeable to a very usable 192GB.

Qnap TS-h1077AFU review: Build quality

The TS-h1077AFU is clothed in the same solid steel chassis as the TS-h1090FU so gets equal praise from us for its build quality. The carriers are the metal tool-free variety and we had no problems fitting SSDs as they are pressed down until they click into place.

The appliance presents a spacious interior with easy access to all components and the Ryzen CPU has a chunky passive heatsink with four DIMM slots to one side. The base 32GB of DDR5 is supplied on one stick allowing memory to be increased to 128GB without having to replace it.

The appliance offers two PCIe Gen4 expansion slots presented on riser cards but we found adapter installation a bit fiddly as each riser must be removed completely before fitting a card. However, if the embedded ports aren't enough, the appliance supports a wide range of Qnap's 10GbE and 25GbE cards.

The comparatively low capacities of SATA SSDs don't make the TS-1077AFU a great choice as a backup vault but you can improve its storage prospects massively by fitting Qnap's SAS3 expansion cards. The appliance supports up to eight JBODs allowing capacity to scale up to 5PB but check system memory requirements as this will probably need upgrading.

Qnap TS-h1077AFU review: Great 10GbE performance

For our 10GbE lab performance tests, we fitted five 1TB Western Digital Red SA500 SATA SSDs and created a single RAID5 array. With a NAS share mapped to a Dell PowerEdge Windows Server 2025 host, we recorded 9.3Gbits/sec and 9.1Gbits/sec for sequential reads and writes with the same results returned for random operations.

IP SAN speeds over 10GbE are just as good with a 500GB iSCSI target returning the same Iometer sequential and random speeds as the NAS share. We ramped up the pressure with a dual 10GbE MPIO link to the target and saw great sequential read and write rates of 18.1Gbits/sec and 18.3Gbits/sec while random operations delivered 18.3Gbits/sec and 17.7Gbits/sec

Swapping to 4K block sizes delivered solid I/O throughputs for the MPIO iSCSI link with Iometer reporting 200,000 and 221,500 IOPS for sequential reads and writes and 200,500 and 149,500 IOPS for random operations. Our random write IOPS test pushed the Ryzen CPU to around 74% utilisation but it never strayed above 17% with all other tests.

Qnap TS-h1077AFU review: QuTS Hero features

The 'h' in its name indicates that QuTS Hero is the preferred operating system for the TS-h1077AFU and for good reason, as it offers superior data protection features to Qnap's more nimble QTS. These include ZFS copy-on-write, end-to-end checksums that handle self-healing of data corruption, and near-unlimited NAS and iSCSI LUN snapshots.

It provides space-saving data compression, and the base memory enables inline deduplication, which requires a minimum of 16GB. Ransomware protection is present with a choice of two WORM (write once, read many) policies that can be applied during NAS share creation.

Backup services are in abundance with the Hybrid Backup Sync (HBS) 3 app managing 3-2-1 protection strategies and securing local appliance data to remote NAS appliances or cloud providers, including Qnap's myQNAPcloud service. The Hyper Data Protector (HDP) for PC/VM app secures VMware and Hyper-V virtualized environments, runs block-level backups of Windows PCs and servers with Qnap's HDP PC agent installed, and the latest version now presents its management console inside the QuTS web interface.

When QuTS Hero h6 is available for the TS-h1077AFU, it will add even more security features. Snapshots get a dedicated management app and enhance ransomware protection with policies for creating immutable snapshots that cannot be tampered with or deleted until the retention period you set has expired.

Other security features on the horizon are the Ransomware Guard which is accessed from the Malware Remover app, detects known malware and monitors processes for suspicious activity. The Secure IP Access feature will augment the QuFirewall app with granular IP-based access controls while the High Availability Manager app allows two identical appliances to be joined in a fault tolerant active/passive cluster.

Qnap TS-h1077AFU review: Is it worth it?

With an MSRP (manufacturer's suggested retail price) of £5,649 including VAT, the TS-h1077AFU will appeal to businesses that want all-Flash performance but without the big price premium for U.2 NVMe SSD support. We expect the actual street price to be lower but if you have slightly deeper pockets, rack space to spare and a desire for a lot more SATA SSDs, you may want to check out Qnap's 2U high, 30-bay TS-h3077AFU with Broadbandbuyer currently listing the 32GB model for £6,819 including VAT.

That said, the TS-h1077AFU offers a powerful hardware package for the price with an impressive SSD density for a 1U rack chassis. It delivers excellent 10GbE performance across the board, the QuTS software presents a wealth of data protection features and the price includes a generous 5-year warranty.

Qnap TS-h1077AFU specifications

Global server market sales hit record levels last year, according to new research from IDC, driven by skyrocketing demand for AI infrastructure.

New figures from the consultancy show organizations continue to ramp up infrastructure spending, particularly with top hyperscalers, while large service providers are racing to lead in the AI era.

As a result, the market saw a staggering 80.4% annual growth rate, recording $444.1 billion in revenue. In the last quarter of 2025 alone, revenue reached $125.3 billion – a 52.4% year-over-year increase.

There was a year-over-year increase of 52.4% in vendor revenue compared with the same quarter of 2024. Revenue generated from x86 servers increased 16.9% in the last quarter to $69.8 billion, while non-x86 servers rose by 146.4% to $55.5 billion.

Elsewhere, revenue for servers with an embedded GPU grew 59.1% year-over-year, representing more than half of the total server market revenue.

Memory price surges are a worry

The rising price of certain components, such as GPUs, DRAM, and SSDs, is causing worry. IDC noted that some enterprises have been trying to deal with the problem by securing prices in advance.

However, the impact could be set to hit harder during 2026 as demand keeps outpacing offering capacity in the near term.

“The race for AI adoption is settling the market pace with companies starving for infrastructure looking not only GPUs but also consuming more CPUs, among other components, in order to feed their needs,” said Juan Seminara, research director of Worldwide Enterprise Infrastructure Trackers.

"We are going to see more price pressures, and that may impact on market dynamics with less units but higher average selling prices going forward."

Dell among the big server market winners

Regionally, the US saw the fastest server market growth, up 72.4% in the fourth quarter of 2024, fueled by 80.1% growth in the accelerated server segment.

Canada grew by 70.7%, for the same reason. EMEA and APeJC, meanwhile, showed double digit growth with 43.6% and 27.9% respectively. PRC and Latin America showed lower growth of 17.7% and 12.8% each, while Japan declined by 4.7%.

Dell Technologies was the clear leader in the OEM market with a 10.0% revenue share, thanks to significant growth on accelerated servers. Supermicro was in second place with 9.5% revenue share for the same reason.

IEIT Systems and Lenovo pretty much tied for third, with 4.1% and 4.0% share respectively, while Hewlett Packard Enterprise finished in fifth place, with a 3.1% share.

FOLLOW US ON SOCIAL MEDIA

As a member of TerraMaster's new Advanced-Level NAS appliance family, the F2-425 Plus delivers a feature set that will appeal to enthusiasts and small businesses alike. At first glance, this compact desktop unit may look like a 2-bay appliance, but it's a hybrid model, as along with dual LFF/SFF front drive bays, it offers three internal M.2 NVMe SSD slots.

The appliance is powered by Intel's 'Twin Lake' Processor N150 CPU, which is a quad-core model with a maximum turbo frequency of 3.6GHz and a very low 6W TDP. The N150 targets a range of applications, including entry-level laptops, embedded systems, and NAS appliances, and is proving to be a popular choice as we've already seen it in mini-PCs such as Beelink's Mini S13 and Acemagic's Vista V1.

The appliance comes with 8GB of fast DDR5 SO-DIMM memory, upgradeable to a very usable 32GB. Multimedia is on the agenda as 4K 60Hz video from the CPU's embedded Intel Graphics GPU is piped through to an HDMI 2.0 port at the rear.

There's a lot more on the table as a peek around the back reveals dual 5GbE multi-Gigabit ports. USB ports are also in abundance, as you have one USB-A 3.2 Gen2 at the front, along with two more at the rear, plus a USB-C 3.2 Gen2, with the appliance supporting TerraMaster's external 10Gbps USB-C drive expansion enclosures.

TerraMaster F2-425 Plus review: Design and build quality

The F2-425 Plus eschews the cheaper plastic chassis of its entry-level brother in favour of an impressively solid 3mm thick aluminium alloy shell. It's a cinch to remove as you undo two cross-head screws in the base and gently slide it off the internal chassis.

The base 8GB of memory comes on a single SO-DIMM stick, which you'll need to lose if you want to upgrade, as there's only one slot. This is surrounded by the three M.2 2280 slots, but bear in mind that the N150 CPU has PCIe Gen3 channels, so don't waste your money on costly PCIe Gen4 NVMe SSDs.

The main SATA LFF drive carriers are the plastic tool-free variety and have mounting holes for SFF SSDs, so you can go all-Flash if you want. Power is provided by a small external 48W adapter, and all cooling is handled by a single 8 cm diameter fan at the rear, which we found to be extremely quiet during testing.

TerraMaster F2-425 Plus review: Deployment and 5GbE performance

For testing, we fitted two Kingston M.2 NVMe SSDs and followed the browser-based quick start wizard to create a mirrored system pool (RAID5 is supported for three SSDs), download the latest TOS 6 software, create a new superadmin user, and register an email address to receive notifications. We also fitted two 4TB Seagate IronWolf NAS drives for general data storage, which were accepted without any problems, as unlike Synology, TerraMaster doesn't have any drive restriction policies.

For 5GbE testing, we have every angle covered as we use Zyxel's XMG1930-30HP multi-Gigabit PoE++ access switch, which offers 24 2.5GbE and four 10GbE ports. Its near-£1,000 price tag is a big ask for small businesses, and although there is a limited choice of low-cost multi-Gigabit switches, you may want to consider TP-Link's TL-SX105 desktop switch, which offers five 1/2.5/5/10GbE ports for around £200.

With a NAS share on the appliance's SSD system pool mapped to a Dell PowerEdge Windows Server 2022 host on our 10GbE LAN, Iometer recorded 4.6Gbits/sec and 3.9Gbits/sec for sequential reads and writes. Swapping to random operations returned read and write speeds of 4.6Gbits/sec and 2.4Gbits/sec.

The appliance supports IP SANs and a 100GB iSCSI target mapped to the server, delivering very similar results to our NAS tests. We increased the pressure with a dual 5GbE MPIO link to the target and recorded respectable speeds of 7Gbits/sec and 5.1Gbits/sec for sequential reads and writes, while random operations delivered 7Gbits/sec and 3.6Gbits/sec.

TerraMaster F2-425 Plus review: Key TOS 6/7 features

The TOS 6 web console is easy on the eye and provides a ribbon menu across the top for quick access to common management functions. Apps are in abundance, with 57 now available for download from the App Center, which also provides a link to a choice selection of community apps.

Data protection features are extensive, with the Backup app providing a single location for quick access to all the suite's associated modules. Along with Rsync support, snapshots of NAS shares and iSCSI LUNs can be run at regular intervals, TFM (TerraMaster folder mirror) Backup handles local or remote copies of shared folders, and CloudSync supports fifteen cloud storage providers, including Amazon S3, Box, OneDrive, and Dropbox.

The Centralized Backup app is similar to Synology's Active Backup for Business, as it can protect physical Windows servers and workstations along with VMware and Hyper-V virtualized environments. We had no problems declaring our Hyper-V and VMware vCenter hosts and could choose which VMs to back up and apply a schedule.

The HyperLock-WORM feature provides valuable ransomware protection as it allows you to assign custom retention periods to volumes during creation. The Enterprise mode allows data to be modified or deleted after the set period has expired, while the Compliance mode stops data from ever being changed – even by administrators.

We also had a sneak peek at the new TOS 7, which is currently in public beta. It adds a heap of new features, along with optionally opening up full root access for developers, and it offers granular access permissions by updating the three basic types to thirteen domain-based ACLs (access control lists) and customisable rules.

TOS 7 presents a redesigned desktop and app icons, adds new Storage Manager and Device Monitoring apps, and increases the number of available apps to 82. The new Online Office is an app of distinction as it allows you to upload MS Office and PDF files to the appliance and edit them directly from the TOS 7 viewer.

TerraMaster F2-425 Plus review: Is it worth it?

Costing a very reasonable £399 including VAT, the F2-425 Plus compares well on value with competing products. Synology's DS725+, for example, costs around £443 but only comes with 4GB of DDR4, two M.2 NVMe slots, and one 2.5GbE multi-Gig port.

The F2-425 Plus is very well built, and its triple M.2 NVMe SSD slots plus dual 5GbE ports add extra levels of versatility. Performance over 5GbE is good, TerraMaster's TOS 6 software offers plenty of useful apps, including a good selection of data protection tools, and the upcoming TOS 7 shows a lot of potential.

TerraMaster F2-425 Plus specifications

Chassis Desktop chassis

Dimensions (WDH) 122 x 219 x 150mm

CPU Quad-core Intel Processor N150

Memory 8GB DDR5 SO-DIMM (max 32GB)

Storage bays 2 x SATA LFF/SFF, 3 x M.2 2280 PCIe Gen 3 NVMe SSD

RAID RAID0, 1, TRAID, 5 (M.2 slots)

Network 2 x 5GbE multi-Gigabit

Other ports USB-C 3.2 Gen2, 3 x USB-A 3.2 Gen2, HDMI 2.0

Power 48W external PSU

Management Web browser

Warranty 2yr limited

It's hard to argue against choosing a Synology NAS for a small business; the brand is the market leader for a reason, and its DiskStation DS725+ is a good example of why. Those familiar with Synology's naming convention will understand this is a '25 model year device. Although it's only a two-bay, the '7' refers to the fact that you can upgrade it with the DX525 five-disk enclosure to give you a much more capable system.

Most other manufacturers let you plug in expansion bays, but here, closer links between the DS725+ and its expansion unit mean you can use the latter to expand an existing disk array in the first. Outgrow the RAID 1 array in your DS725+, and you can add the DX525 with one or more disks and migrate to RAID 5, 6, or 10. That's a great upgrade path for a startup that wants to limit initial expenditure, without having to reinvest if their storage needs scale up.

Elsewhere, the DiskStation DS725+ isn't an especially highly specified device. There's not much wrong with its AMD Ryzen R1600 processor, or the 4GB of RAM you get as standard, but at this price, its single USB 3.2 Gen 1 port is a bit stingy. That's particularly the case given it's chiefly there to connect the DX525 if you need it – once you do, no further expansion is possible.

One other questionable feature comes at the rear, where you'll find the odd pairing of 2.5GbE and 1GbE network ports – many enclosures at this price now offer dual 5GbE. While you can bond both interfaces, we're not sure what their asymmetry means in a multi-user environment – presumably, some transactions benefit from 2.5GbE bandwidth, while some are throttled by 1GbE.

That's it for connectivity; there are no HDMI or audio ports – this NAS can't moonlight as a media centre, and it's not really ideal for surveillance applications.

The DS725+ offers two M.2 slots for NVMe SSDs, which you can use either as a stand-alone fast volume or for read/write caching of the disk array. Here, it's worth noting that Synology currently has the strictest compatibility requirements of any NAS manufacturer. You can use any SATA drive in the main bays, as long as it's not on the incompatibility list, but the operating system will generate a warning for any disk unless it's on the compatibility list. If that seems daft, you can't even use NVMe SSDs not on the list: the DS725+ wouldn't recognise two WD Black modules. Most significantly, at the time of writing, only Synology's own storage products were officially compatible. The company says it's working with third-party manufacturers to broaden this out.

One final thing to mention here is that this enclosure comes with a three-year warranty – a slight advantage when many competitors have only two year backing.

Synology DiskStation DS725+ review: Setting up

As you'd expect, it's physically easy to set up the DiskStation DS725+. It's a small and lightweight device, although it does use a separate power brick. Its lockable drive caddies are tool-less for 3.5" disks, and you get screws in the box for any 2.5" drives. You'll need to flip the enclosure onto its back to access its M.2 slots, which are also tool-free.

This NAS's memory slots are accessed by removing the right-hand drive caddy. It arrives with a single 4GB module and a free slot, making an 8GB upgrade simple and cost-effective. Beyond that, you'll need to remove the supplied module, which supports up to 32GB. While it's ideal to connect and bond both network interfaces, if you're only connecting one, it's important to check that it's the faster of the two.

At nearly five years old, release seven of Synology DiskStation Manager (DSM) is getting a little creaky – version eight is due for release this year. Even so, DSM is still arguably the best NAS operating system out there. Configuration is easy, with simple guides for volume creation at the outset. DSM defaults to Btrf file formats and the added flexibility baked into Synology Hybrid Raid, but you can specify ext4 and standard RAID 1 if you prefer. Helpfully, you can choose to skip a lengthy full drive check for new volumes, while for security, you can encrypt the entire volume from the outset, or shared folders as you create them.

We configured the system with two 6TB WD Red Plus drives in a RAID 1 array. Synology also supplied us with two 800GB SNV3400 series SSDs, which we initially configured as a separate RAID 1 storage pool.

Synology DiskStation DS725+ review: Features and apps

This NAS may not be bristling with physical features, but it makes up for it with a mature operating system and a big collection of apps – 86 at the time of our review. This is where DSM really excels, offering quality apps for a wide range of services beyond simple file serving.

The menu includes DHCP, DNS, mail, proxy, VPN, and Active Directory server apps, for example, alongside open-source resources such as Git Server, Joomla, and WordPress. There are databases, web servers, and surveillance apps, alongside the usual media and streaming services, including Plex. This being 2026, there's also now Synology AI Console, which facilitates external AI integrations with supported Synology apps.

There aren't really any omissions of note, meaning you can configure this enclosure as the hub of a small or growing business. Our only real concern here is that you might need a RAM upgrade relatively soon, especially if you're configuring multiple services alongside data access.

Synology DiskStation DS725+ review: Performance

Those expecting blistering speeds need to remember one simple fact: this NAS' fastest network port tops out at just under 300MB per second. That meant it quite quickly hit the maximum threshold during large file transfers – 5GbE ports would be twice as fast in theory. The odd thing is, however, that the DS725+ didn't seem to get the memo. While it couldn't achieve the impossible, its benchmark performance was a match for 5GbE enclosures over many of our tests.

Even when testing an uncached disk volume, the DS725+ excelled on small file operations, managing 56MB/s write speeds and nearly 60MB/s reads for 4KB operations in ATTO Disk Benchmark. It was mighty on 16KB operations, almost hitting 200MB/s reads. Only on 2MB operations did its lower network bandwidth come into play, with the DS725+ throttled to 296MB/s reads. Unusually, it didn't quite saturate its network connection when writing, achieving 272MB/s. It scored 487 on the PCMark10 Data Drive benchmark, again among the best results we've seen over a 2.5GbE connection. The SSD volume wasn't significantly quicker on benchmarks, with any differences to our disk results mostly fitting within the typical variation between benchmark runs.

Performance on benchmarks is one thing, but happily the DiskStation DS725+ could back it up with strong real-world file copy performance. We copied a mixed 11.1GB dataset to the disk volume from a Windows 11 PC. It took two minutes and 10 seconds, an excellent 92MB/s performance. Reads were even quicker, with the reverse copy complete in two minutes, at 99.2MB/s. The SSDs proved significantly faster here, nudging write speeds up to 105MB/s, and read speeds to 129MB/s – the fastest we've yet recorded on this test.

As with all the NAS devices we test, this enclosure's performance wasn't particularly stellar when reconfigured with the two SSDs on read/write cache duty for the disk volume. We'd expect this could prove faster than a disk-only volume in a multi-user scenario, but on simple transfers it just seems to create overheads, resulting in slightly slower speeds.

Synology DiskStation DS725+ review: Is it worth it?

The Synology DiskStation DS725+ isn't a bad choice for single home workers: it's quite powerful out of the box, and its 2.5GbE connection is more than good enough for typical wireless routers or small network switches. However, it's best suited to startups and micro businesses that need on-site storage that can scale. Here, the ability to add drives and memory should help it keep pace with more users, storing more data, as a small business grows. At the same time, the range of apps on offer provides a low-cost way to add network functionality, potentially delaying the need for investments in more powerful on- or off-premises solutions.

Really, this NAS's only significant weakness is that it has slower network ports than some competitors, with no expansion slot to upgrade them. That may mean it's not ideal for extremely data-intensive use, or for moderate loading in a larger business – although our benchmarks and real-world tests suggest it wouldn't be easily overwhelmed. Provided that's not a concern, it's simply excellent, offering a near-perfect balance between value, performance, and scalability.

Synology DiskStation DS725+ specifications

UGREEN might not necessarily spring to mind if you're looking for a NAS enclosure. Better known for its power accessories, the Chinese manufacturer only officially entered the market in 2024, yet it has a growing offer of desktop devices with up to eight bays, covering everything from casual users through to quite demanding small firms. The UGREEN NASync DXP4800 is its flagship four-bay enclosure, and it may be the best specified we've yet seen.

This is undoubtedly a powerful NAS. It's built around an Intel Pentium Gold 8505 processor, and ships with 8GB of RAM onboard – you can upgrade to a 64GB maximum. Alongside four SATA slots, it offers two M.2 slots for NVMe SSDs. As you'd expect, these can be configured as a fast storage pool or used to cache read and write operations on the disks.

That's not bad at all, but this enclosure's real strength comes in connectivity. The headline here is the rear panel's 10GbE Ethernet port, offering clear daylight over even the 5GbE interfaces beginning to appear on other devices in this bracket. There's a second, 2.5GbE port, which you can bond with the main socket for failover protection and even more throughput. Also at the rear, you'll find two USB2 ports, a USB 3.2 Gen 1 (5Gbps), and an HDMI port supporting 4K video.

On the front, the final highlights are USB 3.2 Gen 2 (10Gbps) USB-C and USB A ports and – unusually – an SD card slot. While the latter is a bit niche, all three could be useful for easy backups and upgrades.

On paper, all this makes the NASync DXP4800 Plus suited to power users and small businesses. With four bays, startups could begin with two drives and add space later as needed. And the inclusion of backward-compatible wired 10GbE Ethernet – rather than fiber – supports those starting out with cheaper, slow equipment, providing a solid upgrade path as needs grow.

UGREEN NASync DXP4800 Plus review: Setting up

It's not just this NAS enclosure's specs that impress. Unpack it, and you'll soon realise it's a nicely designed, well-built bit of kit. The weighty enclosure is metal, and it has a single large cooling fan, rather than two smaller, potentially noisier ones. Pleasing details abound; there's a magnetic metal filter for the fan – you can easily remove it to get rid of any dust. The four drive caddies are lockable and numbered, which helps prevent confusion, and the white status LEDs can display other colors, providing extra information in case of a failure.

As you'd expect, the caddies are tool-free provided you're fitting 3.5" disks. Unusually, there's a clip on the base to extend them widthways: extend, drop the drive in, close the caddy up, and your disk is ready to insert – even less hassle than the standard clip-in rail approach. You'll need to screw in any 2.5" disks, although here the neat design means there's no rail to remove and store when you do so.

You'll need to lay this enclosure on its back to unscrew the cover protecting its M.2 and memory slots – one of those is already occupied with the factory-fitted RAM. Although NVMe speed is unlikely to be the limiting factor on any NAS, it's worth noting that the interface here is PCIe4, rather than the Gen3 more common elsewhere. You get a pair of SSD heat sinks in the box, too.

NAS enclosures aren't often difficult to set up, but UGREEN's UGOS Pro operating system is a joy to use. Give the enclosure a name, set up the admin account, and it's away, quickly configuring the operating system. On restart, the OS guides you through the key tasks of creating a storage pool and shared folder, too.

This drive supports typical four-disk options, including RAID 5, 6, and 10, along with RAID 1 or 0. For this test, we configured four 4TB Synology disks in a 12TB RAID 5 array, and created a 1TB RAID 1 array using two WD Black NVMe SSDs. Unlike some other enclosures, you can't skip synchronisation on a new array; although you can continue to create shared folders and tweak the settings, it's best not to start shifting data around until initialisation is complete, which takes several hours.

Given the painless setup, one oddity is that Windows SMB sharing isn't enabled by default. It's easy to turn it on, just as it's easy to set up features, including load balancing across the two network interfaces. It's also worth noting that this NAS defaults to the EXT4 file format; it also supports Btrfs, though.

UGREEN NASync DXP4800 Plus review: Features and apps

While the hardware is great, and the software makes a strong first impression, it didn't take us long to spot some significant limitations. UGOS Pro is extremely user-friendly, and generally quite capable, but it doesn't have the depth you might expect from, say, Synology Disk Station Manager. This is exacerbated by a comparatively limited App Center, offering only 29 apps at the time of our review.

For business use, the most notable missing apps or features included servers for DHCP, DNS, and a web proxy, but it was a surprise not to see WordPress or Plex, which are almost ubiquitous among smaller NAS products. That said, UGOS Pro does offer both Docker and a proprietary Virtual Machine app, using which users with the patience and know-how can run containerised servers to plug any gaps. There are also UGREEN apps for video and photo management, and DLNA serving.

UGREEN NASync DXP4800 Plus review: Performance

We tested this NAS using a PC with a 5GbE network adaptor, connected via a 10GbE switch. While that meant we couldn't explore its maximum network transfer speeds, it certainly gave us an indication of what power users might experience. It's worth bearing in mind that you'll need fast network hardware to get the most from this enclosure; a 10GbE uplink and 2.5GbE downlinks would particularly help it serve multiple users in a data-intensive small business.

We tested our disk and SSD volumes separately before unmounting the SSD storage pool and reconfiguring the drives as a read/write cache for the disk volume. Using the ATTO disk benchmark gave an overview of direct data performance for various I/O sizes, while we tested real-world performance with a 11.1GB file copy comprising large, medium, and small files – in previous tests, we've found that strong performance on one doesn't necessarily mean a good result on the other.

On ATTO, the UGREEN DXP4800 Plus was one of the fastest NAS enclosures we've yet tested. Whether reading or writing to the disk, SSD, or cached volumes, it managed at least 64 megabytes per second (MB/s) on 4KB operations – a brilliant result. At around 155MB/s its 16KB write speeds weren't quite so exceptional, but at 243MB/s its read performance was astonishing.

As we'd generally expect for 2MB file operations and above, this NAS was throttled by its network bandwidth. It managed a massive 580MB/s or more across the various disk volumes, close to the theoretical maximum bandwidth of 5GbE. Here, it's important to remember that it could probably have gone faster if our test PC had also supported 10GbE.

Sadly, this NAS didn't live up to its potential when it came to Windows file copies. We recorded a fastest write speed of 58.37MB/s, and a read of 75.84: by no means bad, but significantly slower even than some 2.5GbE enclosures we've tested.

We weren't able to load up Plex for our usual transcoding test, but we wouldn't expect this enclosure to struggle with taxing jobs. It's more than highly specified enough to cope with multiple users and concurrent file operations. It's in this multi-access scenario that we'd expect an SSD disk cache to help keep speeds up – it makes little difference in single-user tests.

UGREEN NASync DXP4800 Plus review: Is it worth it?

This is a highly specified and beautifully designed NAS with some very impressive features. It's easy to set up and easy to use, offering lots of storage and performance potential to demanding users. It was also impressively fast on performance benchmarks, if a bit less so in our simple real-world test.

Unfortunately, it's not all good news. UGOS Pro is lacking some of the apps that power users might be looking for, particularly those interested in using the DXP4800 Plus' HDMI port in a media, conference, or surveillance centre application. It's also thin on network services, a potential roadblock if you're looking for one do-it-all box to support your business. That leaves this as a fair choice if you want compact and strongly performing storage for a small business – but only if you won't miss extra features that UGOS Pro currently doesn't offer.

UGREEN NASync DXP4800 Plus specifications

We were mightily impressed with HPE's ProLiant DL345 Gen11 rack server when we reviewed it a couple of years ago. And its successor, the ProLiant Compute DL345 Gen12, builds on its successes. Support for one of AMD's core-heavy Gen5 EPYC 9005 series CPUs keeps the focus firmly on virtualization duties, but it goes further with a doubling of DDR5 memory to 6TB, making it a fine candidate for memory-intensive applications.

HPE is offering an appealing alternative for businesses unhappy with VMware's revamped licensing plans, as the DL345 Gen12 is available with its Morpheus VM Essentials software. Designed to simplify virtualization, its unified management plane can handle its own HVM KVM-based hypervisor clusters along with existing VMware vSphere deployments.

HPE boldly claims its per-socket licensing model can reduce the cost of ownership by up to 90%. You can try it out too, as HPE offers a free non-commercial community edition with a one-year licence for three sockets.

The server's 2U chassis also provides an extensive range of storage features. As you'll see, it can't match the enormous capacities of the DL345 Gen11, but it still has plenty of potential for storage-centric workloads.

HPE ProLiant Compute DL345 Gen12 review: Storage features

There are two specific reasons why storage capacities for the DL345 Gen12 aren't as high as those offered by its predecessor. The DL345 Gen11 has internal midtray and rear drive options, allowing it to support a total of 20 LFF or 32 SSF drives, but these are not available for the DL345 Gen12 – presumably due to the thermal restrictions imposed by the higher TDP Gen5 EPYC CPU.

The good news is the triple-box design at the front of the DL345 Gen12 offers plenty of storage choices with a sharp focus on EDSFF SSDs. Each box supports up to twelve E3.S NVMe SSDs for a total of 36, but the flexible design means you can partner them with standard SFF boxes or a universal media box with two stacked SFF bays and an optical drive.

There are other choices as our system came with one 8-bay SFF box with direct-attached U.3 NVMe SSDs, while its central box sports HPE's NS204i-u V2 boot optimised storage device equipped with a pair of mirrored hot-plug M.2 NVMe SSDs. This multi-purpose box can partner the boot device with two blocks of four EDSFF drives, or you can use one block to present two front-mounted PCIe Gen5 OCP cards.

If it's capacity you're after, the server offers up to twelve SATA/SAS LFF drive bays, or you can sacrifice the top four and replace them with HPE's universal media bay. As with all AMD servers, there is no embedded software RAID, although you can direct-attach SATA devices, which will operate in AHCI mode, and HPE offers plenty of PCIe and OCP 'MR' RAID controllers.

HPE ProLiant Compute DL345 Gen12 review: Remote management

All the latest ProLiant servers come with HPE's iLO7 remote management controller, which is implemented in the DL345 Gen12 as an OCP-compliant DC-SCM (data center secure control module) board. Platform security and supply chain integrity are at the top of its agenda, along with silicon root of trust (RoT) digital firmware fingerprinting, the iLO7 provides protection against next-generation quantum computing threats and HPE's Secure Enclave – a physically tamper-resistant security processor for storing sensitive data.

The iLO7 presents a smart web interface that's been designed around workflows with the customisable home page dashboard presenting up to six cards for at-a-glance status views of key areas such as host health and security issues. Each card has colour-coded icons for each component so you can easily see if anything needs your attention and hot links are provided for drilling down for more information.

We run HPE's OneView in the lab on a Hyper-V host and had no problems importing the DL345 Gen12 for on-premises management and monitoring. The iLO7 also connects directly to HPE's GreenLake Compute Ops Management (COM) cloud service, which provides a complete overview of your server estate plus tools for server deployment and management, automated system lifecycle compliance, and AI-powered insights.

HPE ProLiant Compute DL345 Gen12 review: Build, cooling and expansion

Cracking the lid reveals a very tidy internal design and, unlike the Xeon-powered DL340 Gen12 we recently reviewed, all the cabling is neatly arranged and secured out of harm's way. The 64-core 3.3GHz AMD EPYC 9575F CPU in our air-cooled system is looked after by a high-performance heatsink with radiator extension bars.

Lower specified systems can use HPE's standard cooling fans with our system, including a full house of six hot-plug high-performance fan modules. HPE offers its DLC (direct liquid cooling) kit for this server, but not its nifty closed-loop liquid cooling system, which is currently only available for the DL320, DL325, DL36,0 and DL560.

Expansion potential is excellent as there's room at the rear for two three-slot risers that support full-height, half-length PCIe Gen5 cards. Lurking beneath are two OCP 3 PCIe Gen5 slots, which support HPE's 'o' RAID cards and a good selection of network adapters, with our system kitted out with HPE's quad-port Gigabit and dual-port 25GbE modules.

The DL345 Gen12 expands GPU support further over the Gen11 model as it can take six 75W single-width (SW) or four 400W double-width (DW) cards. HPE's front enablement kit provides four or two full-height, full-length PCIe 5 slots for DW GPUs and there's space in between them for a central box for EDSFF or SFF storage devices.

HPE ProLiant DL345 Gen12 review: Is it worth it?

The core-dense Gen5 EPYC CPU and huge memory capacity makes the ProLiant Compute DL345 Gen12 a very affordable solution for virtualization hosting and many other demanding enterprise workloads. It won't beat the previous Gen11 model for storage capacity, but it does deliver a highly flexible range of options, brings HPE's smart iLO7 management into play and has a remarkable expansion potential with a high GPU support.

HPE ProLiant DL345 Gen12 specifications

Chassis 2U rack

CPU 64-core 3.3GHz AMD EPYC 9575F

Memory 768GB 6,400MT/s ECC DDR5 Smart Memory (max 6TB)

Storage bays 8 x hot-swap SFF (max 24 with three boxes)

RAID None included

Storage included 4 x 3.2TB direct-attach U.3 NVMe SSDs

Cooling 6 x high performance hot-plug fans

Network 4 x Gigabit OCP, 2 x 25GbE OCP

Expansion Up to 6 x PCIe Gen5 slots, 2 x OCP 3 slots

Power 2 x 2,400W Titanium hot-plug PSUs

Management HPE iLO7, OneView, GreenLake COM

Warranty 3Yrs On-Site NBD

It isn't often we get to see a completely new rack server from HPE, but the Compute DL340 Gen12 is its first ProLiant model to deliver a single socket (1P) Intel Xeon-powered solution in a 2U rack chassis. This market space has traditionally been held by HPE's DL345 rack servers, as their AMD EPYC CPUs have the core densities to make them good alternatives to more costly two-socket (2P) servers.

The DL340 Gen12 targets a wide range of data centre and service provider duties, and the high core counts of Intel's Xeon 6 CPUs allow it to offer a balanced mix of performance, storage capacities, energy efficiency. And, of course, costs. It supports plenty of Xeon 6 models, so you can choose from P-Core CPUs up to 86 cores and E-Core CPUs up to 144 cores, or pick from all six P-Core 'Rich I/O' models with up to 80 cores that are specifically optimised for 1P deployments.

Memory capacity is good, as regardless of the CPU model chosen, the server can present up to 4TB of HPE's DDR5 Smart Memory – only its speed is determined by the CPU. That extra 1U of chassis height brings a lot of storage permutations into play too, as it can handle a diverse range of LFF, SFF, and EDSFF storage devices.

HPE simplifies the acquisition process as it offers three CTO (configure to order) options where you take the base model, add a CPU and memory, and select a drive backplane. The third CTO option has AI workloads firmly on its radar as it supports up to four front-mounted double-width (DW) or single-width (SW) GPUs.

HPE ProLiant Compute DL340 Gen12 review: Build quality and cooling

As we've come to expect from HPE, chassis build quality is excellent, and good internal design offers easy access for upgrade and maintenance tasks. Our only complaint is that the two M-XIO cables, along with the power and signal connector cables for the NS204i-u boot device, are excessively long and look untidy.

The CPU socket sits towards the front and is flanked on each side by banks of eight DIMM slots. HPE supplied our system with a 32-core 2.5GHz Xeon 6730P CPU topped off with an impressively large high-performance heatsink complete with two large radiator bars that extend forward and sit close to the front cooling fans.

A key feature of the DL340 Gen12 is air cooling is supported for all specifications, so you don't need to worry about the extra expense of liquid cooling. You have two fan choices, and the 512GB of memory in our system dictates that we need the six hot-plug high-performance fans.

HPE ProLiant Compute DL340 Gen12 review: Storage choices

The DL340 Gen12 scores highly in the storage department as it employs a three-box design at the front with each one supporting a variety of configurations. For LFF hard disks, the three, quad-drive boxes are arranged horizontally with the top one also supporting two SFF devices and an optical drive.

SFF configurations use different boxes with vertically mounted devices and you can have up to 24 hot-plug HDDs and SSDs or use the first one for two stacked drive bays and an optical drive. It gets even more interesting for EDSFF drives as each box supports up to twelve E3.S NVMe SSDs, and you can mix and match these with the standard SFF boxes.

The DL340 Gen12 scores over AMD servers which don't have embedded RAID controllers. Its motherboard sports Intel's Virtual RAID on CPU (VROC) for HPE ProLiant Servers embedded, which is specifically for up to 32 NVMe SSDs with direct to CPU connections and supports software-managed stripes, mirrors, and RAID5 arrays.

For other storage devices, there are HPE's 'MR' hardware RAID controllers and choices abound as HPE offers three OCP and two PCIe versions. Our system was fitted with the rear-mounted MR408i-o OCP 3.0 card with 4GB of energy pack protected cache, plus support for SATA, SAS4, and NVMe device,s and all the usual RAID array suspects.

There are plenty of placement options for HPE's NS204i-u V2 boot optimised storage device, which can be fitted at the front or at the rear above the PSU bays or next to them. It comes with two hot-plug 480GB or 960GB M.2 NVMe SSDs preconfigured in a mirror for running an OS or hypervisor.

HPE ProLiant Compute DL340 Gen12 review: iLO7 management

Remote management features are outstanding with HPE's new iLO7 controller implemented as an OCP-compliant DC-SCM (data center secure control module) board. Platform security and supply chain integrity are assured as the iLO7 silicon root of trust (RoT) digital firmware fingerprinting is future-proofed against next-generation quantum computing threats, and it is endowed with HPE's patented Secure Enclave - a physically tamper-resistant security processor for storing sensitive data such as encryption keys, passwords, and security configurations.

The slick new iLO7 web console is designed around workflows and presents a customisable home page dashboard with up to six cards. Colour-coded icons show details such as host health and security issues, and each one provides hot links so you can quickly drill down for more information

We run HPE's OneView on a Hyper-V host for on-premises monitoring of our HPE systems and had no problems importing the server's iLO7 into its console. Next up is HPE's GreenLake Compute Ops Management (COM) cloud service, which connects directly to the server's iLO7 controller and provides a portal for viewing your entire infrastructure, managing server deployments, automating system lifecycle compliance policies, and loading AI-powered insights.

HPE ProLiant Compute DL340 Gen12 review: Is it worth it?

The ProLiant Compute DL340 Gen12 is a versatile rack server that will appeal to businesses seeking a more cost-effective and energy-efficient alternative to expensive dual-socket Xeon 6 servers. It offers a high CPU core density and big memory capacity, the 2U chassis presents impressive storage options and expansion potential, while HPE's iLO7 controller delivers outstanding platform security and remote server management services.

HPE ProLiant Compute DL340 Gen12 specifications

Global server market growth skyrocketed in the third quarter of 2025, new research shows, fueled by continued enterprise cloud and AI spending.

Figures from IDC’s Worldwide Quarterly Server Tracker show the market hit a record-breaking $112.4 billion in revenue across Q3.

This, the consultancy noted, marked yet another double-digit growth rate, with the industry recording 61% growth compared to the same period last year.

Revenue generated from x86 servers increased by 32.8% to $76.3 billion, according to IDC. However, the biggest growth rate came in Non-x86 servers, hitting $36.2 billion and marking a huge 192.7% YoY increase.

Notably, revenue for servers with an embedded GPU now accounts for more than half of broader server market revenue, with growth in this domain surging 49.4% YoY.

IDC said a key factor here lies in sharpened hyperscaler focus on AI infrastructure projects, with major providers ramping up hardware spending to accommodate rapid enterprise adoption rates.

All told, spending on this front has helped fuel broader server market growth over the last 18 months, with IDC noting that the market has doubled in size compared to 2024, reaching revenues of $314.2 billion in the first three quarters.

“IDC expects AI adoption to keep growing at an outstanding pace as major vendors continue reporting record orders and showing strong backlogs,” said Juan Seminara, research director, Worldwide Enterprise Infrastructure Trackers.

“Hyperscalers and cloud providers are still ahead with new, large deployments that require much higher compute density. Additionally, we started to see major AI based Research and Education projects that will help fuel further growth.”

US market surge continues

IDC’s tracker shows continued growth across a host of markets, with the US in particular the fastest-growing region due to heightened hyperscaler infrastructure spending.

Indeed, the US recorded a 79.1% increase compared to the third quarter of 2024. IDC added that a key contributing factor here was a 105.% growth in accelerated server spending.

Market growth in China also surged by 37.6% YoY, with the region accounting for nearly one-fifth of global quarterly revenue.

“APeJC, EMEA and Japan also showed very healthy double digit growth with 37,4%, 31.0% and 28,1% respectively,” IDC added. “While Latin America showed a low single digit growth with 4.1% increase in the quarter.”

The big boom time winners

Major vendor positions in the server market have fluctuated over the last year, IDC found. Dell Technologies holds a solid lead, boasting an 8.3% revenue share which the consultancy attributed to "outstanding growth on accelerated servers”.

Supermicro, meanwhile, holds second place with a 4% share. However, IDC noted that its position here compared to last year has slipped by 13.2%.

Elsewhere, Lenovo and IEIT Systems remain tied in third position, commanding a 3.6% and 3.7% share respectively.

MORE FROM ITPRO

• Dedicated servers are back in vogue as IT leaders scramble to meet AI, compliance requirements
• Best business servers
• How to set up a server for small businesses

With their fire-engine red chassis, there's no mistaking WatchGuard's Firebox security appliances, and the M695 on review here comes in at the top of the company's freshly launched product range. Stepping up as a replacement for the four-year-old Firebox M690, it claims a substantial speed increase and teams this up with a versatile range of high-speed network ports.

Targeting large environments with around 1,250 users, the M695 moves to Intel power as this 1U rack appliance replaces the elderly NXP LX2160A SoC (system on chip) in the M690 with a more modern Gen14 8-core 2.6GHz Intel Core i7-14701E CPU. System memory sees a doubling to 32GB of DDR4 while internal storage is handled by a 128GB M.2 SSD.

The new hardware delivers an impressive performance, boasting a high raw firewall throughput of 45Gbps and 10.2Gbps with antivirus (AV), the intrusion prevention service (IPS), and application controls all activated. Compared with the M690, these numbers represent increases of 51.5% and 122% respectively.

WatchGuard Firebox M695 review: Hardware features

A quick glance at the front of the M695 shows that WatchGuard has removed the port expansion slot from the previous series. Instead, you get a much wider range of fixed ports that should cover most connection requirements.

Across the front, you have banks of eight and four 2.5GbE multi-Gig ports, four Gigabit SFP fibre, four 10GbE SFP+ fibre, and two 10GbE RJ-45 copper ports. One feature absent is PoE support, as the M690 accepted WatchGuard's optional four-port with PoE+ expansion module. Although we always thought this was a bit of a kludge, as it requires an extra 54V power brick plugged into a dedicated port at the rear to enable power delivery.

Power protection sees changes as all the new M-series appliances, except the entry-level M295, have a single PSU at the rear with an expansion bay next door. This accepts WatchGuard's proprietary WG9037 250W redundant PSU module, which costs around £340.

WatchGuard Firebox M695 review: Simple subscriptions

WatchGuard's subscription schemes cut through any confusion as you only have to choose from two. All Firebox appliances with a support contract come as standard with firewall, VPN, software-defined WAN (SD-WAN), and clientless VPN access portal services enabled.

WatchGuard's Basic Security Suite (BSS) subscription enables gateway antivirus (GAV), web filtering, antispam, HTTPS inspection, IPS, application controls, network discovery, and WatchGuard's RED (reputation enabled defence) cloud-based URL filtering. A Total Security Suite (TSS) subscription activates everything WatchGuard has to offer and includes its advanced persistent threat (APT) blocker with cloud sandboxing and ThreatSync XDR, which is part of WatchGuard's Automation Core (WAC) and provides collection, correlation, plus automated responses for threat events from all Fireboxes.

There's more as DNSWatch monitors client DNS requests and blocks access to known malicious domains, while the IntelligentAV anti-malware service uses the Cylance AI-based engine to scan files such as Office documents, Windows portable executables, and PDFs after they've passed through the GAV scanner. Access to the WatchGuard cloud portal for remote monitoring and management is included in both subscriptions, with TSS increasing its log and report retention periods to one year and 30 days, respectively.

WatchGuard Firebox M695 review: Cloud management

Deployment for cloud management is impressively swift as we registered the appliance with our WatchGuard support account, allocated it to our site, and chose the cloud management and monitoring option. Once the M695 had been configured, local management access was disabled, with its local web console only providing details of device and network status, options for firmware upgrades, and access to diagnostics tools

Another bonus of cloud management is zero-touch provisioning using WatchGuard's RapidDeploy service. Create a configuration file, upload it to your cloud account, send the appliance to the remote site, and once connected to the internet, it automatically downloads and applies the file for instant protection.

The portal dashboard tells you everything you need to know about all your Fireboxes, their status, and licence details, along with alerts and pending incidents. All other WatchGuard products can be accessed from the same portal and include its Wi-Fi access points, AuthPoint, EPDR, FireCloud, and the AI-powered ThreatSync + NDR service.

WatchGuard Firebox M695 review: Security services

Firebox configuration doesn't get any easier as all appliance hardware and security settings are presented in one web page. The mixed routing mode makes the M695 highly flexible, as you can segment your network by configuring each port as separate interfaces each with its own IP address and DHCP services, and designate them as external, trusted, optional, or custom.

Providing an alias for each port also makes it easier to assign custom firewall rules to them. Note that the network discovery service is not supported by cloud-managed Fireboxes as it can only be enabled in their local web console.

The content scanning section provides access to both AV scanners, the APT blocker, and spamBlocker for applying anti-spam policies to inbound SMTP, IMAP, and POP3 traffic. The network blocking section looks after botnet detection, IPS, port and site blocks, and detection of Tor (The onion router) exit points.

Move to the content filtering section, and you can create WebBlocker policies to manage access to 169 URL categories. While the Application Control service provides 1,267 predefined app and protocol signatures for creating granular access policies. Both can be applied to a firewall policy with one click.

VPN features are plentiful, with the M695 supporting 2,000 branch offices and mobile VPNs. WatchGuard's free mobile VPN with SSL client is available for Windows and macOS systems, and the latest Fireware firmware supports strong SAML user authentication.

WatchGuard Firebox M695 review: Is it worth it?

The Firebox M695 compares favourably on value with the competition in the enterprise gateway security space. WatchGuard Online is listing the appliance with a 1-year TSS subscription for £19,282 excluding VAT, rising to £35,184 for a 3-year subscription, making it very competitively priced.

The new Intel-powered architecture delivers a significant performance boost over its predecessor, and the appliance offers a good selection of high-speed network ports as standard. It's packed with great security features, the simple licensing schemes avoid any confusion, and WatchGuard's slick cloud portal makes light work of remote deployment, management, and monitoring.

WatchGuard Firebox M695 specifications

HPE's ProLiant Compute DL325 Gen12 offers businesses a great alternative to more costly dual-socket (2P) rack servers. Supporting most of AMD's Gen5 EPYC CPUs, this single-socket (1P) server can squeeze up to 192 cores into only 1U of rack space.

Targeting a wide range of enterprise workloads, including hybrid cloud, containerized environments, and virtualization, it takes everything we liked from its predecessor, the ProLiant DL325 Gen11, and makes it even better. For starters, DDR5 memory capacity has been doubled to an impressive 6TB, and the DL325 Gen12 sports HPE's new iLO7 remote management controller.

Storage capacities remain the same as the DL325 Gen12 also supports up to four LFF, ten SFF, or 20 E3.S EDSFF NVMe SSDs, but the server is available with HPE's new multi-purpose drive bays, allowing you to mix and match different storage devices. HPE's NS204i-u V2 hot-pluggable boot-optimized storage device can now be mounted at the front or the rear, as can the OCP expansion slots.

HPE ProLiant DL325 Gen12 review: Build quality and cooling

Chassis built quality is excellent, and the server presents a tidy internal design offering easy access for upgrade and maintenance tasks. Our system was supplied with a 32-core 3.55GHz EPYC 9355P CPU mounted centrally, topped off with a large passive heat sink and flanked on each side by banks of 12 DIMM slots.

The server currently supports 23 different EPYC models ranging from 8 up to 192 cores, but choose your CPU carefully as there are significant cooling requirements. Those with TDPs up to 240W can use the standard heatsink and cooling fans, while models between 260W and 300W (which includes our CPU) need the performance versions, with all configurations requiring a full house of seven fans.

For CPUs with TDPs of 320W and upwards, you must specify liquid cooling, and the good news is the server supports HPE's nifty closed-loop liquid cooling system. We've seen this in action when we reviewed the DL325 Gen11 and found the kit takes up very little room as the CPU cold plate with integral pump has two coolant pipes routed directly to a full-width radiator mounted between the cooling fans and the drive backplane.

HPE recommends replacing the liquid cooling kit every five years, which is now field replaceable. The new iLO7 management controller also introduces a coolant leak detection feature, which wasn't available in the iLO6.

HPE ProLiant Compute DL325 Gen12 review: Storage and expansion

The chassis may only measure a shade under 67cms deep, but HPE has packed it full of storage and expansion goodness. Our review system was supplied with eight SFF drive bays, a single tri-mode backplane that supports SATA, SAS4 plus NVMe SSD devices, and a blanking plate to the right that can be replaced with an extra dual-drive SFF cage or a DisplayPort/USB/optical drive enablement kit.

Other storage configurations are a four LFF drive backplane or the hybrid storage solution, which comprises up to five drive boxes, with each one supporting two SFF drives or four EDSFF E3.S NVMe SSDs. The GPU-optimized chassis supports up to four 75W single-width or two 400W double-width GPUs, leaving room at the front for two drive boxes with either four SFF drives or eight EDSFF E3.S SSDs.

AMD's chipset doesn't offer any integral RAID features, so you'll need to specify one of HPE's Smart Array Gen11 (there are currently no Gen12 models) controller cards, which are available in PCIe or OCP versions. We didn't need a RAID card in our system as it had four 3.84TB NVMe SSDs with the backplane direct-attached to the motherboard's M-XIO connectors.

The multi-purpose drive bays also support front-mounted OCP NIC cards and HPE's NS204i-u V2 boot-optimized storage device. We had the latter installed at the rear and fitted with two hot-plug 480GB M.2 NVMe SSDs to provide mirrored redundant storage for running an OS or hypervisor.

There's plenty of room to grow as the server has two rear risers, each offering PCIe Gen5 x16 expansion slots. Underneath lurk two OCP 3 PCIe 5 slots, which support HPE's 'o' RAID cards and a good range of network adapters.

HPE ProLiant Compute DL325 Gen12 review: iLO7 management

Remote management sees big changes as HPE's new iLO7 controller delivers tougher platform security and supply chain integrity. Furthermore, unlike the DL360 Gen12 we recently reviewed, the iLO7 is implemented in this server as an OCP-compliant DC-SCM (data center secure control module) board.

The iLO7 silicon root of trust (RoT) digital firmware fingerprinting is future-proofed to protect against next-generation quantum computing threats, and it introduces HPE's patented Secure Enclave. This is a physically tamper-resistant security processor that stores sensitive data such as encryption keys, passwords, and security configurations.

The new iLO7 web console is designed around workflows and offers a customizable home page dashboard that can present up to six cards. These have color-coded icons showing details such as security issues and server health, and each one provides hot links so you can quickly drill down to see more information

We run HPE's OneView 10.2 in the lab on a Hyper-V host and had no problems importing the DL325 Gen12 for on-premises management and monitoring. The iLO7 also connects directly to HPE's GreenLake Compute Ops Management (COM) cloud service which provides a complete overview of your server estate plus tools for server deployment and management, automated system lifecycle compliance and AI-powered insights.

HPE ProLiant Compute DL325 Gen12 review: Is it worth it?

Businesses balking at the cost and power demands of dual-socket rack servers will find HPE's ProLiant Compute DL325 Gen12 offers a very cost-effective alternative. The top TDP EPYCs will require liquid cooling, but this single-socket server combines a high core density with an impressive memory capacity, teams them up with a versatile range of storage features, adds HPE's highly secure iLO7 controller, and delivers it all in a surprisingly small package.

HPE ProLiant Compute DL325 Gen12 specifications (as reviewed)

Businesses seeking a powerful rack-optimised server will find HPE's ProLiant Compute DL360 Gen12 could be just what they're looking for. Supporting up to 144-core Xeon 6 CPUs, this dual-socket 1U server targets a diverse range of high-performance workloads such as server consolidation, hybrid cloud, databases, VDI, data analytics, and virtualization.

The Gen12 portfolio comprises eight servers and introduces a wealth of new features, including enhanced platform security. HPE claims they are the world's first to meet the FIPS 140-3 Level 3 certification and will be resistant against future quantum computer attacks.

The family is designed to offer businesses the right size for their workloads, and the DL360 Gen12 will appeal to those who want rack-dense power combined with a high memory capacity, improved scalability, and plenty of storage options. The DL360 Gen12 also has AI workloads on its radar as it supports up to three single-width (SW) GPUs.

HPE bags another industry first with the introduction of a 100% fanless direct liquid cooling (DLC) solution, which comprises an 8-element design that extends this to the CPUs, GPUs, local storage, and network fabric. Smaller sites can still opt for air-cooled models or go for HPE's innovative closed-loop cooling system, which impressed us when we reviewed its ProLiant DL560 Gen11 and DL325 Gen11 models.

HPE ProLiant Compute DL360 Gen12 review: Under new management

All ProLiant Gen12 servers come with HPE's latest iLO7 remote management controller, which delivers an impressive range of new features. Supply chain integrity and platform security are at the top of the list, as the silicon root of trust (RoT) digital firmware fingerprinting has been future-proofed to protect against next-generation quantum computing threats.

Leak detection for liquid-cooling wasn't available in the Gen11 servers but the iLO7 rectifies this and introduces HPE's patented Secure Enclave. Embedded in the iLO ASIC, this is an independent, physically tamper-resistant security processor that stores and protects highly sensitive data such as encryption keys, passwords, and security configurations.

HPE is also making moves to embrace the DC-MHS (data center modular hardware system) specification by implementing the iLO7 as an OCP-compliant edge-connected board – but not on all Gen12 servers. The DL320 Gen12 and DL340 Gen12, for example, use DC-SCM (data center secure control module) boards whereas our DL360 Gen12 review sample has the iLO7 controller embedded on its motherboard.

The iLO7 presents a smart new web interface that's been designed around workflows, with the customisable home page dashboard presenting up to six cards for at-a-glance status views of key areas such as host health and security issues. The cards have colour-coded icons for each component so you can easily see if anything needs your attention, and hot links are provided for drilling down for more information.

We run HPE's OneView on a Hyper-V host for on-premises monitoring of all our HPE systems, and after updating it to the latest 10.2 version, we were able to import the server's iLO7 into its console. Next up is HPE's GreenLake Compute Ops Management (COM) cloud service, which connects directly to the server's iLO7 controller and presents a web portal for viewing your entire infrastructure, managing server deployments, and automating system lifecycle compliance policies with this latest version, adding AI-powered insights and improved map-based visibility.

HPE ProLiant Compute DL360 Gen12 review: System design

The server presents a well-designed interior providing easy access for upgrade and maintenance tasks. We were supplied with the air-cooled model, which has seven dual-rotor hot-plug HPC fans mounted behind the drive backplane.

The server supports two Xeon 6 P-Core CPUs up to 86 cores and E-Core models up to 144 cores. Our system was supplied with a pair of 32-core 2.5GHz Intel Xeon 6730P CPUs topped off with chunky passive heatsinks sporting front radiator extensions.

Each CPU is flanked on each side by sixteen DIMM slots and if you specify P-core CPUs, you can use 256GB modules and push total capacity to a whopping 8TB. E-core CPUs support maximum DIMM sizes of 128GB and drop capacity to 4TB.

There's plenty of room for further expansion as the server has two risers at the rear offering up to three PCIe Gen5 x16 slots. Underneath these are two OCP 3 PCIe Gen5 slots which support HPE's 'o' RAID cards and a good range of network adapters with our system supplied with HPE's quad-port 10GbE module.

HPE ProLiant DL360 Gen12 review: Storage features

The server excels in the storage department as the chassis is available with four LFF or up to ten SFF hot-swap drive bays. Of more interest is the hybrid storage solution, which came with our review system.

Up to five drive boxes can be installed at the front, with each one supporting two SFF drives or four EDSFF E3.S NVMe SSDs. The backplanes for the SFF boxes are the TriMode variety, which extend support to 24G SAS drives and NVMe SSDs.

The motherboard has Intel's Virtual RAID on CPU (VROC) for HPE ProLiant Servers embedded, which is specifically for NVMe SSDs with direct to CPU connections and supports software-managed stripes, mirrors, and RAID5 arrays. For other storage devices, there are HPE's hardware RAID controllers with a choice of four OCP and two PCIe versions.

The bays are all multi-purpose as they can support storage devices, an enablement kit for front-mounted OCP NIC cards, and HPE's NS204i-u V2 boot optimised storage device. We had the latter fitted in the fifth front bay, which comes with two hot-plug 480GB M.2 NVMe SSDs and provides mirrored redundant storage for running an OS or hypervisor – you can also fit it at the rear if you want to free up the front storage bay.

HPE ProLiant Compute DL360 Gen12 review: Is it worth it?

Very much so, and HPE has made the ordering process even easier with the new Smart Templates. These represent the top-selling configure-to-order (CTO) models and can assist in building a system based on the customer's workload requirements.

HPE's ProLiant Compute DL360 Gen12 packs an impressive hardware package and is well-suited to rack-dense deployments that demand plenty of power. Along with three cooling choices, it combines flexible storage options with a high expansion potential, and HPE's new iLO7 delivers great management features and reassuringly strong platform and supply chain security.

HPE ProLiant Compute DL360 Gen12 specifications

Chassis 1U rack

CPU 2 x 32-core 2.5GHz Intel Xeon 6730P

Memory 1TB DDR5 HPE SmartMemory (max 8TB)

Storage bays 2 x hybrid drive boxes (max 5)

RAID Embedded Intel (VROC) for HPE ProLiant Servers

Storage included 2 x 1.6TB SFF NVMe SSDs, 4 x 1.6TB EDSFF E3.S NVMe SSDs, HPE Front NS204i-u V2 with 2 x 480GB NVMe SSDs

Network 4 x 10GbE OCP 3

Expansion 3 x PCIe Gen5, 2 x OCP 3 Gen5 slots

Power 2 x 2,200W 80Plus Titanium hot-plug PSUs

Management HPE iLO7 Standard

Warranty 3Yr On-Site NBD

Remote server management is a must-have feature in all data centres, and Dell's iDRAC (integrated Dell Remote Access Controller) has always been one of our favourite solutions. Included as standard across Dell's Gen17 enterprise server family, the iDRAC10 delivers an impressive range of capabilities, and the latest firmware upgrade introduces a wealth of new features, including a radically revamped web interface.

A major hardware change is that whereas previous PowerEdge generations have the iDRAC controllers embedded on their motherboards, the iDRAC10 is deployed as a DC-SCM (datacenter-ready secure control module). The DC-SCM moves all server management and security functions to a dedicated OCP-compliant edge-connected board, allowing Dell to use a standard form-factor across the entire Gen17 range and develop these functions independently of the server.

Supply chain assurance and access security see major improvements as the iDRAC10 controller sports a dedicated security processor with a cyber-resilient architecture. It supports MFA (multi-factor authentication), integrated Root-of-Trust (RoT) encryption, and device-level attestation using stronger algorithms to verify the integrity of all hardware components during the secure boot phase.

Dell iDRAC10 review: Design

Introduced over eight years ago, the iDRAC9 controller hasn't seen any significant changes to its web interface for a long time. With firmware version 1.20.60.55 applied, the new iDRAC10 web interface is a very welcome update as it isn't just a refresh, but a complete redesign.

The iDRAC9 home page presents a basic overview of the server's health status, along with the most recent logs, access to the virtual console, and a ribbon menu across the top for accessing all other functions and services. The iDRAC10 version is far more informative as the new server health information section in the home page presents a group of active icons at the top showing the health status of all critical hardware components, and these are all color-coded for at-a-glance alerts.

You still get a server summary, a list of the most recent log entries, and access to the virtual console, but all menu options have been moved to a side menu for easier access. The firmware update removes AngularJS to address known security vulnerabilities and enables support for Intel's ME CUPS (Compute Usage Per Second) feature on all Gen17 Intel PowerEdge servers, providing real-time monitoring of CPU, memory, I/O, and system utilization.

The changes aren't skin-deep either, as moving through the various menus shows their entire layout has been updated with the aim of providing a more consistent view. It's easier to navigate, too, as each menu in the left pane expands downwards to show all available options.

Selecting the system menu extends out a set of sub-menus below, with the overview page providing a row of tabs across the top to quickly view system information. The new environments tab provides one screen with a detailed view of cooling, temperatures, and fan status, along with a historical graph of system inlet temperatures going back a day, month or year.

As we were using a PowerEdge R7725 server for testing, there are (still) no CPU, memory, and I/O usage graphs in the system menu but this is entirely due to AMD's architecture and is a problem Dell has been tussling with for years. No such issues with Intel-based PowerEdge servers and CUPS support allows the iDRAC10 to present performance data on all these components with data samples taken every six seconds and aggregated over one-minute intervals.

The main storage menu now only has two sub-menus with the overview page providing tabs across the top for viewing controllers, physical disks, virtual disks, and enclosures. A slightly earlier firmware upgrade adds support for Dell's PERC12 RAID controllers and the PERC13 H975i Front card.

The new interface also makes it easier to access details presented in tables. In iDRAC9, you click the plus sign next to an item whereas with iDRAC10, you choose an item, such as an entry in the job queue, and click on the double-arrow icon to the right to expand it in an adjacent side panel.

Dell iDRAC10 review: Performance

The iDRAC10 sees significant hardware improvements over its predecessor as the CPU is updated from a 32-bit dual-core ARM CPU to a quad-core 64-bit model and is AI-ready with its ARM Neon SIMD architecture and floating point units. The DC-SCM board keeps the standard dedicated Gigabit network port for out-of-band server management and you'll also find a USB-C port at the front of all Gen17 PowerEdge servers for walk-up direct access to the iDRAC web console for secure emergency maintenance.

Dell's SCV (secured component verification) provides digital proof to partners and VARs that the server they receive matches what was produced at the factory. The iDRAC10 update adds a new encrypted delta certificate process to further verify hardware authenticity.

Management gets a boost as Dell has expanded Redfish integration, as the new firmware supports the latest Schemas and Specifications, with the Dell developer portal providing access to the Redfish APIs. This allows businesses to integrate server management and monitoring with their existing solutions – we run Paessler's PRTG in the lab and were able to add its Redfish sensors to gather detailed information on system health, power supplies, and storage.

Dell's OpenManage Enterprise (OME) software provides centralized discovery, monitoring, and remote control for all its servers, and we updated our instance to version 4.5 to add support for iDRAC10 controllers. Its interface also sees improvements as once we'd applied an OME Enterprise Advanced licence to our server's iDRAC10, it enabled the Power Manager plug-in, which now directly accessed from the OME home page ribbon menu

The OME free version includes discovery, hardware reporting, and health monitoring. All OME versions support the AIOps Observability plug-in, allowing them to function as collectors and send server telemetry and alert logs to Dell's cloud-hosted management platform.

Dell iDRAC10 review: Is it worth it?

Of course it is. Dell has also simplified licensing with iDRAC10 Core included with all Gen17 servers, and providing essential system management services. An Enterprise licence enables virtual media services and full OS remote control, while a Datacenter licence adds thermal management and telemetry streaming of hardware metrics for advanced analytics.

Along with big improvements to the web interface, this latest iDRAC10 update delivers tighter platform and supply chain security plus advanced monitoring and management services. The update is now available for all Dell customers, so if you are already using Gen17 PowerEdge servers, what are you waiting for?

Dell iDRAC 10 specifications

Every server vendor has a flagship product, and for Dell, it's always been the PowerEdge R7x0 series, as these 2U rack-mounters have traditionally been its most capable models that can turn their hand to just about any workload. The PowerEdge R770 on review takes this to the next level as it combines powerful Xeon 6 processing with vast amounts of DDR5 memory, offers plenty of storage permutations, and teams it all up with a highly flexible, modular design.

Typical workloads targeted by the R770 include virtualization, hyperscale, software-defined storage (SDS), web services, and scale-out databases. It also has AI/ML duties high on its agenda as it supports a wide choice of Nvidia and Intel GPUs, along with fast NVMe SSD storage and high-bandwidth networking options.

It can't match the sheer EPYC core density of the PowerEdge R7725 we recently reviewed, but the R770 supports two Xeon 6 P-Core CPUs up to 86 cores and E-Core models up to 144 cores. As with the R7725, it is available with air-cooling or direct liquid cooling (DLC) kits, but is more versatile as it is offered in hot and cold aisle chassis configurations.

Dell PowerEdge R770 review: Design and expansion

There's a lot of hardware inside this server, but Dell's excellent internal design makes everything very accessible. In common with all 2U Gen17 PowerEdge servers, the R770 sports a T-shaped motherboard that allows the PSUs to be positioned on each side of the chassis to improve airflow.

Our system was supplied with a pair of 64-core 2.4GHz Intel Xeon 6767P CPUs, and cooling for their 350W TDPs is ably assisted by impressively large passive heatsinks with additional radiators fore and aft. This arrangement allows the system to be specified with six hot-plug high-performance (HPC) Silver fan modules, which worked well in practice as the CPUs stayed as cool as cucumbers during testing.

Each CPU is flanked on each side by sixteen DIMM slots, and if you specify P-core CPUs, you can use 256GB modules and push total capacity to an incredible 8TB, whereas E-core CPUs support maximum DIMM sizes of 64GB and drop capacity to 2TB. Dell will also be offering the new CXL (compute express link) 2.0 memory for P-core equipped models.

Expansion potential is quite remarkable, as you have three rear risers, each offering two PCIe Gen 5 slots. These are all easily removed by undoing six thumbscrews, and underneath, you'll find two carriers, each with a single PCIe Gen 5 slot.

There's more, as below the carriers are two OCP 3.0 slots with another connector in the centre for Dell's BOSS-N1 DC-MHS (datacenter modular hardware system) adapter. The right-hand OCP slot in our system was occupied by the iDRAC10 DC-SCM (datacenter-secure control module), with the other home to a Broadcom quad-port 25GbE network card – Dell also offers 1GbE, 10GbE, 100GbE, and 400GbE OCP options.

Dell PowerEdge R770 review: System management

Dell is clearly moving away from proprietary solutions for some server hardware components, and implementing the iDRAC10 remote management controller as a DC-SCM has many advantages. Placing all server management and security functions on a dedicated OCP-compliant board allows Dell to use a standard form factor across the entire Gen17 range and develop new features independently.

And new features are in abundance – the iDRAC10 has a dedicated security processor with a keen focus on platform and supply chain security as it enforces integrated Root-of-Trust (RoT) encryption, offers MFA (multi-factor authentication), and verifies hardware components with stronger algorithms during its secure boot phase. The iDRAC10 web console doesn't see any significant changes, but none are really needed as it's clearly the one to beat for features.

Dell estates will approve of the OpenManage Enterprise (OME) software, which provides centralised discovery, monitoring, and remote control for all its servers. We run version 4.4 as a Hyper-V VM in the lab and had no problems importing the R770's iDRAC10 into its console.

The OME free version includes discovery, hardware reporting, health monitoring, and firmware updates. An advanced licence enables the Power Manager plug-in along with server template deployment and configuration compliance, while the Advanced+ licence brings in ServiceNow integration and the VMware OMEVV plug-in. All OME versions support the APEX AIOps Observability plug-in, where they function as collectors and send server telemetry and alert logs to Dell's cloud-hosted management platform every five minutes for servers with a valid support contract.

Dell PowerEdge R770 review: Storage options

Storage sees significant changes over the R760 as the R770 isn't offered with LFF hard disk bays. That's no big deal if you're after sheer storage capacity, as Dell's lower-cost single-socket Xeon 6 R570 may be more appropriate.

The R770 is available with a great choice of SFF options ranging from 8 up to 24 SAS4/SATA devices, and if cooling is a priority, you can drop to 16 and have Dell's Smart Flow panel in the centre. The system is also available with universal drive slots so you can add up to 8 NVMe SSDs and have them direct-attached or managed by one of Dell's PERC RAID cards.

Support for E3.S NVMe SSDs is excellent as you can start with 8 devices and go up to 40 in the hot-aisle chassis. Choose the cold-aisle version and you can slot up to 16 in between the front-facing I/O panels.

More changes are afoot with Dell's fPERC RAID cards as along with the BOSS-N1 controller, they are all now DC-MHS compliant. Developed under the OCP initiative, DC-MHS is designed to standardise on server hardware components and reduce vendor lock-in, and Dell claims to be the first server OEM to fully embrace this initiative.

Dell PowerEdge R770 review: Is it worth it?

Starter prices look good as £6,300 excluding VAT gets you an entry-level R770 Smart Selection model equipped with dual 16-core 2.3GHz Xeon 6515P CPUs, 64GB of DDR5, an fPERC H965i, 16 SAS4/SATA drive bays with Smart Flow panel, one 960GB SATA SSD, two 1,100W Titanium PSUs and a 3-year ProSupport NBD contract. The system supplied to us for review costs nearly £57K excluding VAT which is comparatively good value as it delivers a very powerful package for the price.

The PowerEdge R770 is probably Dell's most versatile server yet as it combines Xeon 6 power with a huge memory capacity, plenty of GPU choices and a wide range of storage options. Available in hot and cold aisle configurations, it offers an impressive expansion potential and Dell's wholehearted adoption of the DC-MHS specification is to be applauded.

Dell PowerEdge R770 specifications

Organizations are increasingly reliant on dedicated servers, new research from Liquid Web shows, and AI performance requirements are a key factor in their revival.

In a survey of more than 1,000 IT professionals across industries and company sizes, 86% said their organization is currently using dedicated servers, with government (93%), information technology (91%), and finance (90%) the most likely to do so.

The trend appears even within smaller businesses, with 68% of micro companies reporting using them - and these numbers are growing. In the past 12 months, 42% of respondents said they've migrated workloads from public cloud back to dedicated servers.

Just over one-third of respondents (34%) said their organizations increased spending on dedicated servers in 2024 - more than in any previous year.

"This upward trend in 2024 spending reflects a broader shift back to dedicated environments, with 42% of IT professionals saying they migrated workloads away from the public cloud in the past 12 months," Liquid Web said.

In terms of workloads, they're being mainly used for databases, large-scale file storage, web hosting, internal developer environments, compliance-sensitive systems, and AI model training or inference.

Just over half of respondents said they'd never use public cloud for regulated data such as healthcare or finance, with others citing core databases, performance-intensive computing, and AI workloads.

“People outside of tech often think servers are outdated or only for legacy systems. But dedicated infrastructure still plays a critical role in compliance-heavy industries, and sometimes outperforms cloud for specific workloads," said one survey respondent.

“The biggest misconception is that we don’t need them anymore because of cloud adoption,” another respondent said. “But dedicated servers still handle our most critical operations due to reliability, control, and predictable costs.”

IT leaders eye more control with dedicated servers

More than half the IT professionals (55%) said their main reason for choosing dedicated infrastructure over cloud was full control and customization. Other reasons included network performance, physical security, predictable pricing, and custom hardware configurations.

Unfortunately, some said they faced communication challenges within their organizations, with nearly a third saying they don’t feel confident explaining infrastructure choices to non-technical leadership figures.

More than half of IT professionals (53%) said they still consider dedicated servers essential, and nearly as many (45%) believe their importance will grow even more by 2030.

However, there’s still room for improvement, with a quarter of respondents asking for easier scaling, with others looking for access to modern developer tools (15%) and sustainable energy use (14%).

"Despite predictions of cloud dominance, dedicated servers remain a foundational component of enterprise infrastructure, especially in industries where performance, control, and compliance are critical," said the researchers.

"IT professionals continue to rely on them not out of nostalgia, but because they deliver predictable, customizable environments that serve mission-critical needs."

MORE FROM ITPRO

• AI security and compliance concerns are driving a private cloud boom
• Where next for private cloud networks?
• There’s a ‘cloud reset’ underway, and VMware Cloud Foundation 9.0 is a chance for Broadcom to pounce on it

Enterprises looking to put some serious processing power into their data centers will find the PowerEdge R7725 blows everything else away. Topping Dell's Gen17 server portfolio, this dual-socket (2P) rack mounter supports AMD's EPYC 9005 CPUs right up to the bad boy 192-core 9965 – that's an incredible 384 physical cores in only 2U of rack space.

It's no surprise the R7725 is targeting demanding workloads such as big data analytics, AI/ML, HPC and virtualization but it will also appeal to businesses with big server consolidation plans. Flexibility is the name of the game as along with support for 6TB of 6,400MT/s DDR5 memory, the R7725 has room for a wide range of storage configurations including up to forty E3.S Gen5 NVMe SSDs.

Expansion potential is equally impressive as the server offers up to eight PCIe Gen5 slots and teams them up with dual OCP 3 slots. AI workloads are covered as it can handle two 400W double-width (DW) or eight 75W single-width (SW) GPU cards and you can specify air-cooled systems with Dell's DLC (direct liquid cooling) kits available for higher-spec systems.

Dell PowerEdge R7725 review: Under the hood

Removing the lid reveals a superbly designed interior with the server employing Dell's standard T-shaped motherboard. This allows the PSUs to be located on each side of the chassis to create a more balanced air flow which ensures there are no hot spots on one side of the server.

General chassis cooling is handled by a bank of up to six hot-plug fan modules behind the drive backplane. We have a full house of Gold high-performance fans but Dell also offers Silver and Platinum versions.

Dell clearly wanted to show off to us as it supplied the R7725 review system with dual 192-core EPYC 9965 CPUs. We also had the DLC kit fitted with the pipework from the CPU's low-profile cold plates incorporating leak detection systems and neatly routed through the center of the chassis to the rear.

The R7725 will appeal to businesses that aren't ready for liquid cooling in their data centres as Dell's Smart Cooling technology is so efficient it allows the server to support most configurations in air-cooled environments. The 9965 CPUs may have a hefty 500W TDP but as long as you're mindful of the thermal restrictions on GPU use, the R7725 can be air cooled if you specify T-wing heatsinks and Platinum fans.

Dell PowerEdge R7725 review: Storage

The R7725 is highly flexible in the storage department as it supports a maximum of 12 LFF or 24 SFF SAS/SATA drives. Unlike the R7625, it doesn't support rear drive cages but you can mix and match 16 SFF SAS/SATA drives with 8 U.2 NVMe SSDs.

AMD chipsets don't have any embedded RAID features but that's no problem as you can choose from three of Dell's PERC RAID controllers. These range from the entry-level front-facing fPERC H365i right up to the latest high-performance PERC13 H975i which has a sharp focus on storage acceleration for AI workloads.

The system on review was supplied with two 16-bay E3.S cages with the backplanes cabled directly to the motherboard's embedded PCIe signal connectors. In between them is a blank panel with a grill for improved airflow and this can be replaced with another drive cage for a total of 40 E3.S NVMe SSDs.

You can keep all your drives for data storage as the R7725 supports Dell's BOSS-N1 DC-MHS (data center modular hardware system) adapter. Slotting in at the rear, it provides dual hot-plug M.2 NVMe SSD slots with a choice of two 480GB or 960GB sticks which can be mirrored or striped.

Dell PowerEdge R7725 review: Server management

Dell's iDRAC10 is a standard setter for remote management and introduces a wealth of new features including a dedicated security processor endowed with a cyber-resilient architecture. It offers MFA (multi-factor authentication), integrated Root-of-Trust (RoT) encryption and stronger algorithms to verify hardware component integrity during the secure boot phase.

Gen17 PowerEdge servers have the iDRAC10 implemented as a DC-SCM (datacenter-ready secure control module) which moves server management and security functions off the motherboard and onto an OCP-compliant edge-connected board. It's a smart move as Dell can develop new features independently of the server and use a standard form-factor across all the range.

The iDRAC10 web interface presents a wealth of easily accessible information about server operations, component status and health along with a full hardware inventory and direct access to BIOS and RAID configurations. You can configure alerts for pretty much any event you want and we were pleased to see the iDRAC10 had options to create these for the DLC system and apply actions such as a graceful power down or immediate shutdown if coolant leaks are detected.

We run Dell's OpenManage Enterprise (OME) software as a Hyper-V VM in the lab which provides discovery, monitoring, remote control and power management for all its servers. You'll need this if you want to use Dell' s APEX AIOps Observability cloud hosted infrastructure management platform as OME functions as a collector and sends server telemetry and alert logs to it every five minutes.

Dell PowerEdge R7725 review: Is it worth it?

Prices start at around £13K excluding VAT for entry-level models but you'll need deep pockets if our system takes your fancy as this comes in at over £105K although the 20 E3.S NVMe SSDs account for nearly half of this. That said, the PowerEdge R7725 will pay back handsomely on your investment as it offers the industry's highest core density in a 2U rack server.

It's a highly versatile server well-suited to a wide range of enterprise workloads and is ideal for businesses aiming to replace their underperforming legacy servers with fewer systems and free up valuable rack space. Storage and expansion potential are very good, management and monitoring services are beyond reproach and Dell's smart cooling design allows it to run most specifications in air-cooled environments.

Dell PowerEdge R7725 specifications

TP-Link's Omada SX3832MPP is geared up for SMBs looking to make the move to high-performance Wi-Fi 6 and Wi-Fi 7 networks. Stepping up as the flagship member of TP-Link's Access Max series, this 1U 32-port Power over Ethernet (PoE) switch has what it takes to handle the latest access points (APs) as it offers 24 10GbE multi-Gigabit ports each capable of delivering up to 90W of 802.3bt PoE++ power.

It can easily cope with a high density of powered devices (PDs) as it delivers a generous PoE budget of 770W. It's also aimed at businesses that want to build a complete 10GbE network as the last eight ports are the 10GbE SFP+ variety for high-speed uplinks over long distance fibre.

At its foundation, the SX3832MPP is a Layer 2 (L2) switch with basic Layer 3 (L3) routing capabilities. Often referred to as 'L2+' or 'Layer 3 Lite', it supports static IPv4 and IPv6 routing, but not the dynamic routing found in more expensive full Layer 3 switches.

Management options are in abundance as smaller businesses can deploy the switch in standalone mode and use its local web interface. Those planning on hanging other TP-Link devices off the switch such as its APs, will prefer cloud management and there are a wide range of choices available.

TP-Link Omada SX3832MPP review: Hardware features

The switch is solidly constructed with an all-steel chassis and employs four cooling fans in the left-hand side which are audible but hardly intrusive. It also incorporates 6kV surge protection circuitry making it a good choice for powering outdoor PDs such as APs and IP cameras.

The SX3832MPP has the capacity to handle a high data throughput as it's endowed with a speedy 640Gbits/sec backplane. Combine this with a high 240 million packets per second (Mpps) forwarding rate and you have a switch that easily has the measure of more expensive 10GbE SMB products.

Visual operational cues are plentiful as along with status LEDs for power and system, you have one that warns if you're reaching the maximum power budget and another for fan failure alerts. Each port has two LEDs showing the negotiated speed and whether PoE delivery is active. Should you wish to delve into its CLI (command line interface), the switch offers USB-C and RJ45 serial ports.

TP-Link Omada SX3832MPP review: Standalone mode

In standalone mode, the web console sensibly requests the default administrative password to be changed on first contact and then opens with a system overview. This presents a graphic of the switch showing which ports are active and moving to the PoE view reveals how much of the budget is in use, which ports are supplying power and the amount in Watts.

A menu across the top provides easy access to all L2 and L3 features with the latter offering options to create static routes plus DHCP servers and relays. L2 features are in abundance and include port, MAC and protocol-based VLANs, QoS traffic prioritisation, port mirroring plus static and LACP link aggregation groups.

VoIP networks are covered as the switch identifies traffic from IP phones using their OUI (organizationally unique identifier) and automatically prioritizes it by dynamically creating voice VLANs. Security features are plentiful too, and include 802.1X port-based authentication, DoS defenses, and access control lists.

The big power budget makes some of the switch's PoE features redundant, although if you're worried about exceeding it, you can assign priorities to individual multi-gig ports. Assign one of three priorities to each one and if the drain reaches the power threshold, those with the lowest priorities will be switched off first.

There's a nice surprise for switch monitoring as along with support for all SNMP versions, the switch has an internal sFlow agent for sending packet sampling data to a collector device. We pointed the agent at our lab host running the WhatsUp Gold 2024 network monitoring software which reported back on switch port utilization, traffic throughout, the top detected apps, and all endpoint conversations.

TP-Link Omada SX3832MPP review: Cloud management

TP-Link offers plenty of cloud management choices with hardware and free software on-site controller appliances plus cloud-based versions. We opted for the latter and tested the free Omada Essentials cloud controller which provides a base set of management and monitoring features for small businesses.

We run multiple Omada cloud controllers which can be selected from the portal's organisation view. Choosing our Essentials controller took us to its own portal which offers a dashboard view of all managed devices but, unlike the Standard version, this can't be customised.

We added the switch to our site using the Omada iOS app to scan its barcode and selecting it from the portal's device page provided a graphic to the right showing active ports, their connection speed, and those providing power. An overview below reveals the available power budget and tabs are provided for assigning profiles to each port, rebooting powered devices, and viewing attached clients.

One drawback of cloud management is that a few features such as port PoE priorities and the sFlow agent can't be accessed from the Essentials or Standard portals. The switch's local web interface is disabled when the cloud controller manages it, so you'll need to configure these features in standalone mode first and apply a cloud profile that instructs the switch to keep its existing settings after adoption.

Omada's new IntelliRecover feature will appeal to support staff. Currently in beta phase, it can be used to monitor powered devices and automatically reboot them if they fail to respond.

TP-Link Omada SX3832MPP review: Is it worth it?

Costing around £1,800 excluding VAT, TP-Link's Omada SX3832MPP delivers an impressive range of features for the price along with plenty of management choices. It teams up a high density of 10GbE multi-Gig ports with full PoE++ services and its generous 770W power budget makes it a great choice for SMBs deploying high-performance Wi-Fi 6/7 services.

TP-Link Omada SX3832MPP specifications

Nvidia CEO Jensen Huang joined Michael Dell on stage today during the opening keynote of Dell Technologies World 2025 as the two CEOs discussed their broadening partnership on Dell AI Factory with Nvidia.

The two men first joined forces on AI Factories last year in what is fast becoming one of the key trends in how enterprises use AI and is where Nvidia in particular is putting a lot of its efforts.

The announcements cover software and hardware, as well as managed services.

Expanded Dell PowerEdge line with Nvidia

Six new servers were introduced at the event, including the air-cooled PowerEdge XE9780 and XE9785, and the liquid-cooled XE9780L and XE9785L, which all support up to 192 Nvidia Blackwell Ultra GPUs with direct to chip cooling. They can also be customized with up to 256 Nvidia Blackwell Ultra GPUs per Dell IR7000 rack and, the company claims, deliver up to four-times faster large language model training than their predecessor, the PowerEdge XE9680.

Another new product is the PowerEdge XE9712 with Nvidia GB300 NVL72, which, Dell Technologies claims, offers 50 times more AI reasoning inferencing output. The new Dell PowerEdge XE7745, meanwhile, will offer the Nvidia RTX Pro 6000 Blackwell Server Edition when it's released in July 2025.

"This is a great addition to the AI factory with Nvidia that Dell has, because it's going to really help customers – especially enterprise customers – adopt AI servers faster than ever before," said Varun Chharba, SVP of infrastructure and telecom marketing at Dell. "The unique capabilities of the RTX Pro 6000 help organizations harness the full potential of agentic AI. They help enable digital twins to revolutionize existing workflows."

Updates to AI Data Platform and agentic software

Further hardware updates to elements of Dell AI Data Platform were also announced including:

• Integration of Nvidia BlueField-3 and Spectrum-4 to Dell ObjectScale
• A high-performance solution built using DellPowerScale, Dell LProject Lightning and PowerEdge XE servers, integrating Nvidia NIXL Libraries
• Nvidia AI Enterprise software platform is now available directly from Dell
• Red Hat OpenShift is available on Dell AI Factory with Nvidia.

Finally, the companies announced Dell Managed Services for the Dell AI Factory with Nvidia, which Dell Technologies claims will simplify AI operations with the management of the full Nvidia AI stack.

Dell Technologies World 2025 runs from 19-22 May. Catch up with all the latest news and analysis live from the event with our live blog.

“The future is flash”, according to Willi Song, president at Huawei European Enterprise Business, and the tech giant appears all in as it looks to streamline enterprise AI adoption.

Speaking at the Innovative Data Infrastructure (IDI) Forum in Munich, Germany, Song and colleagues detailed the company’s sharpened focus on flash storage, unveiling the official launch of its AI Data Lake and OceanStor Pacific all-flash series.

A key factor behind this push, attendees were told, lies in the benefits flash can offer enterprises in the age of AI. It’s more energy efficient, scalable, and cost effective compared to traditional storage solutions, and industry providers have been keen to highlight these advantages in recent years as firms grapple with rising costs.

At Pure Accelerate 2024, the annual conference hosted by Pure Storage, the company waxed lyrical about the enterprise benefits of flash, while competitors such as NetApp have been ramping up similar offerings on this front.

Indeed, research by SNS Insider shows the all-flash array market is booming, reaching a value of $14.59 billion in 2023 and expected to reach over $74 billion by 2032.

Tapping into this lucrative market is a huge opportunity for Huawei, but it does face stiff competition - it’s butting heads with heavyweights like Dell Technologies, HPE, and the aforementioned duo.

Huawei agility a key appeal

Peter Zhou, president of Huawei Data Storage Product Line, told ITPro one key differentiator for the company lies in its significant R&D investment..

“We’re very strong on R&D investment,” he said. “Some data storage vendors, they’ve got less investment than Huawei - because the data storage market was occupied by public cloud. But what we’ve been seeing is that the importance of data will be emphasized more and more.”

Huawei currently operates more than 29 R&D centres across Europe, along with two open labs and boasts close collaborative ties with over 200 universities. The company has made great efforts to embed itself across a variety of industries in the region and is very much playing the long game.

In his opening keynote, Song noted the company’s commitment to European operations.

“Europe is one of our most important markets,” he said. “We will stay true to our ‘in Europe for Europe’ strategy and strengthen collaboration with all sectors in the region.”

Zhou insisted that Huawei has been highly consistent on its European ambitions and stressed it’s something that enterprise customers appreciate. Moreover, the fact it’s had to muscle in on an already crowded market means it knows customers crave reliability and speed and can deliver on both

“This kind of consistency is different,” he told ITPro. “Because Huawei is not seeking short-term wins in the market. Our investment is looking to long-term strategy.”

“We always try to have a very quick response for our customers. My R&D team are not only in China, they’re also in Munich, in Switzerland. Some of my R&D team are sitting with the sales team, and the purpose is to try and listen into the requirements - even the complaints - from our customers and then make a response very quickly.”

Intel's huge Xeon 6 family takes the fight right to AMD's doorstep, and Dell's latest Gen17 PowerEdge servers aim to take full advantage of these super-massive CPUs. The new PowerEdge range currently comprises four models, and in this hands-on review, we bring you a first look at the PowerEdge R570.

Targeting medium to large workloads, the R570 offers a balanced mix of features and capacity. Its single-socket (1P) design will appeal to businesses seeking an affordable, energy-efficient solution ideally suited to virtualisation, medium VM density, or VDI and SDS (software-defined storage) node duties.

CPU choices are extensive as the R570 has room for an Xeon 6 E-Core up to 144 cores or P-Core up to 86 cores, with the latter supporting up to 4TB of DDR5 memory. The R570 offers plenty of GPU options and introduces Dell's new iDRAC10 remote management controller, while its 2U high chassis offers an impressive expansion potential and has a lot to say in the storage department with support for up to 32 EDSFF E3.S NVMe SSDs.

Dell PowerEdge R570 review: System design

A new chassis design makes the R570 highly versatile as it's available in cold-aisle and hot-aisle configurations. The former option reduces the maximum number of EDSFF SSD bays to sixteen but will appeal to support staff as it presents all expansion slots and I/O ports at the front for easier maintenance and servicing access.

We were supplied with the hot-aisle chassis, and cracking its lid reveals a remarkably tidy interior affording easy access to all key components. The server employs a T-shaped motherboard, which allows the PSUs to be placed on each side of the chassis to improve air flow throughout and ensure there are no hot spots.

Our system was blessed with a mighty 86-core 2GHz Xeon 6787P CPU sitting underneath a very large passive heatsink and flanked on each side by eight DIMM slots. As we had a P-Core CPU, memory can be expanded up to 4TB, whereas an E-Core CPU reduces this to 1TB.

Chassis cooling is handled by a bank of up to six hot-plug fan modules in front of the motherboard. Dell has clearly done a fine job with air flow, as even with the 86-core CPU and 1.5TB of memory in residence, our system only required five HPR Silver fans – more powerful models will need up to six HPR Gold fans.

Expansion options are impressive, as further back are a set of stacked risers offering up to four PCIe Gen5 slots, while underneath are two OCP 3 Gen5 slots. GPU options are equally good, as there's room for up to four 75W single-width Nvidia cards with support for three 400W double-width cards on the future feature list, although you'll need to be mindful of Dell's thermal restrictions when fitting these.

Dell PowerEdge R570 review: Storage abundance

The server gets top marks for storage features as the chassis supports up to 12 LFF and 24 SFF SAS/SATA devices, with the hot-aisle chassis having space for 32 EDSFF E3.S Gen5 NVMe SSDs. Our system had two 8-bay SFF front cages with room for a third 8-bay cage in the middle, and if you're prepared to lose the right-hand expansion riser, you can add a four-bay E3.S SSD cage at the rear.

The R570 won't be supporting the next-gen PERC 13 controllers, but RAID options are still plentiful with our system sporting Dell's 'value performance' PERC 12 H365i Front (fPERC) adapter, which nestles in a dedicated slot behind the drive backplane. It supports SAS4, SATA, and NVMe devices but lacks cache memory, whereas the 'premium performance' H965i card includes 8GB of DDR4 cache memory. If you want to expand storage outside the box, Dell offers the HBA 465e and H965e external adapter cards.

Dell's BOSS-N1 DC-MHS (data center modular hardware system) adapter is supported, and in our system, it was fitted in a slot at the rear – in a cold-aisle chassis, it can be presented at the front for easier access. The card provides dual hot-plug M.2 NVMe SSD slots with a choice of two 480GB or 960GB sticks, and we used the latter to create a mirrored array and deploy Windows Server 2025 to it.

Dell PowerEdge R570 review: Remote management

Dell's iDRAC is a standard setter for remote management, and the R570 introduces the new iDRAC10, which has a sharp focus on security. New features include a dedicated security processor with a cyber-resilient architecture, MFA (multi-factor authentication), integrated Root-of-Trust (RoT) encryption, and device-level attestation using stronger algorithms to verify the integrity of all hardware components during the secure boot phase.

A major hardware change is that whereas previous PowerEdge generations have the iDRAC controllers embedded on their motherboards, the iDRAC10 is now deployed as a DC-SCM (datacenter-ready secure control module). The DC-SCM moves all server management and security functions to a dedicated OCP-compliant edge-connected board, allowing Dell to use a standard form-factor across the entire Gen17 range and develop these functions independently of the server.

Plenty more management services are available with Dell's OpenManage Enterprise (OME) software, providing discovery, monitoring, and remote control for all its servers. We run OME in the lab as a Hyper-V VM and, after updating it to version 4.3.2, could import the R570's iDRAC10 into its console.

Adding an OME Enterprise Advanced licence to the server's iDRAC10 enables the Power Manager plug-in, which presents global power consumption and thermal stats for all monitored servers. For individual servers, you can view graphs showing CPU, storage, memory, and fan power consumption plus airflow in CFM (cubic feet per minute) and enforce per-rack power cap and thermal event-based policies.

With a valid support contract, you can integrate OME with Dell's APEX AIOps Observability cloud hosted platform. OME functions as a collector for AIOps, sending telemetry and alert logs to it every five minutes, and you can use the cloud portal to keep a close eye on your entire Dell infrastructure, view PowerEdge server status and system performance, and use the Cybersecurity feature to monitor their security posture.

Dell PowerEdge R570 review: Is it worth it?

The PowerEdge R570 is a highly versatile 1P, 2U rack server, and with Intel's Xeon 6 CPU at the helm, is a great choice for businesses with big virtualisation and storage plans that want to avoid the expense of a 2P server. Build quality is excellent, it offers a wide range of storage configurations with plenty of room to grow, and the new iDRAC10 delivers outstanding remote server monitoring and management services.

Dell PowerEdge R570 specifications

Qnap's TS-h765eU will appeal to space-poor SMBs as this 1U rack NAS is a mere 292mm deep. That means it'll easily fit in a small wall box, a standard two-post data cabinet, or unobtrusively on a desk. It joins Qnap's burgeoning family of low-profile rack NAS models and stands out thanks to its innovative storage arrangements. Up front you have a standard set of four hot-swap carriers supporting SATA LFF/SFF HDDs and SSDs and a peek around the back reveals three E1.S form factor bays.

The appliance includes three E1.S trays in the base price and these are already fitted with M.2 2280 NVMe SSD carriers. Alternatively, you can install higher-capacity E1.S (short) PCIe NVMe SSDs and the appliance supports 5.9mm, 9.5mm, and 15mm models.

The TS-h765eU includes dual 2.5GbE multi-Gigabit network ports but if you feel the need for more speed, the E1.S slots have another role to play. The SSD carriers can be swapped out for Qnap's optional QXG-ES10G1T E1.S to 10GBase-T network modules with up to two supported.

Qnap TS-h765eU review: Hardware features

The TS-h765eU is the most powerful of Qnap's short-depth appliances as it sports a comparatively modern quad-core 1.7GHz Intel Atom x7405C SoC (system on chip) with burst speeds of up to 3.4GHz and a miniscule 12W TDP. It supports DDR5 memory although the base 8GB can only be expanded to 16GB and with one SO-DIMM slot on the motherboard, you'll have to replace the existing module.

The appliance doesn't include a mounting kit with Qnap's optional Rail-S01 sliding rack rail kit costing around £66. You can expand storage outside the box as the appliance supports Qnap's 4-bay TR-004U USB 3.2 1U rack shelf which is the same depth as the TS-h765eU.

The E1.S trays are easily removed and to fit M.2 NVMe SSDs, you slot them into the carrier and let the retaining tab snap back to secure them. For E1.S PCIe NVMe SSDs, it depends on the type you choose as for 5.9mm SSDs, you simply replace the M.2 adapter and screw them directly onto the trays.

For thicker 9.5mm and 15mm models, you need to purchase Qnap's optional Tray-E1S-01 carriers and don't waste your money on PCIe 4 devices as the Atom CPU only has PCIe 3 lanes. Qnap offers optional heatsinks for M.2 and E1.S SSDs and we recommend using them as after creating a cache with two 800GB M.2 NVMe SSDs, we noticed the appliance's three cooling fans ramping up – the appliance is almost silent when only using HDDs.

Qnap TS-h765eU review: Deployment and software features

We loaded four 22TB Western Digital Red Plus NAS-specific HDDs, used the browser-based quick start wizard to load the latest QuTS hero OS, and then created a big 58TB RAID5 storage pool. You can install Qnap's more nimble QTS software but QuTS offers a lot more data protection features such as ZFS copy-on-write, near unlimited NAS and iSCSI LUN snapshots, and end-to-end checksums that handle self-healing of data corruption.

Ransomware protection is available as QuTS offers a choice of two WORM (write once read many) policies that can be applied during NAS share creation. QuTS also offers deduplication but this feature is only available when you have 16GB of memory.

QuTS offers plenty of backup apps with the Hybrid Backup Sync (HBS) 3 managing 3-2-1 protection strategies and securing local appliance data to remote NAS appliances or cloud providers including Qnap's myQNAPcloud service. The Hyper Data Protector (HDP) app secures VMware and Hyper-V virtualized environments, presents a dedicated management web console, and can run block-level backups of Windows PCs and servers that have Qnap's free NetBak PC agent installed.

Security services are abundant and include the Antivirus, Malware Remover, and QuFirewall apps. The Security Center app uses predefined scan policies to search for and draw your attention to vulnerabilities and it includes an unusual file activity monitor.

Qnap TS-h765eU review: 10GbE performance

For performance testing, we installed two QXG-ES10G1T modules. With a NAS share mapped over 10GbE to a Dell PowerEdge R760xs Windows Server 2022 host, we recorded Iometer sequential read and write rates both of 9.2Gbits/sec and swapping to random operations returned 9.2Gbits/sec and 9Gbits/sec.

IP SAN speeds are on the money too, with a 1TB target delivering the same speeds over 10GbE as our NAS share. We ramped up the pressure with a dual 10GbE MPIO link to the target and watched sequential read and write speeds increase to 18.4Gbits/sec and 12.9Gbits/sec. Changing to random read and write operations delivered 18.4Gbit/sec and 10.9Gbits/sec and although these numbers are impressive, we did see CPU utilization peaking as high as 95% during the write test.

It's worth noting that we didn't use an SSD cache during all our NAS and IP SAN tests and the Iometer random tests show clearly how efficient the QuTS ZFS ARC (adaptive read cache) and ZIL (ZFS intent log) are. A cache may provide some extra speed benefits but if you're installing expensive E1.S NVMe SSDs, it would make more sense to use them to present a standard high-performance storage pool.

Qnap TS-h765eU review: Is it worth it?

With Qnap's UK online store listing the TS-h765eU at £790 excluding VAT, it looks to be well worth it. Despite its miniscule dimensions, the appliance delivers an impressive hardware package and the clever storage design and 10GbE upgrade options make it very versatile.

Bearing in mind the Atom x7405C CPU supports up to 32GB of DDR5 memory, it's a shame Qnap has limited expansion to 16GB – especially as the preferred QuTS hero OS is more memory-hungry than QTS. The good news is the appliance delivered a superb performance in our 10GbE lab tests even with only 8GB onboard.

The TS-h765eU is a great choice for SMBs seeking a powerful NAS appliance that won't take up much space. Qnap also offers extra peace of mind as it is committed to making the TS-h765eU available and supported up to 2031.

Qnap TS-h765eU review specifications

Microsoft has resolved an issue that saw companies running Windows Server 2022 and Server 2019 unexpectedly upgrading to Windows Server 2025 — despite not paying for the upgrade.

Windows Server 2025 was released earlier this month with new features including upgrades to security in Active Directory, hotpatching for hybrid and on-prem clouds, and more. However, it also came with a few notable bugs, including one that crashed computers.

Shortly after the release, users spotted that older versions of Windows Server were suddenly upgrading to Windows Server 2025.

One employee of a small business in the UK posted on Reddit that Windows 2022 Servers had upgraded themselves to the new version, or were about to do so.

"This obviously came as a shock as we're not at the point to do so for many reasons and the required licensing would not be present," the user wrote.

Unwanted Windows Server 2025 upgrades

Microsoft said in an update that there were two scenarios in which this happened. The first involved devices automatically upgrading to the new OS.

"This was observed in environments that use third-party products to manage the update of clients and servers," Microsoft explained.

The tech giant noted the issue had been mitigated, but also advised users to ensure third-party software designed to manage updates be configured to not automatically deploy them.

Microsoft updated the release under the upgrade classification "optional" — suggesting there was an issue with how the software was classified for admins and third-party software.

That was backed up in a separate post by a technical manager at security and update software company Heimdal, which noted that the "KB" label appeared to be incorrect. Whether that caused the unexpected updates or not, Microsoft has fixed the "KB" issue.

"The Windows Server 2025 feature update was made generally available on November 1, 2024, as KB5044284, which was the same KB number used for Windows 11, version 24H2," Microsoft said in its update.

"This was the KB numbering for both these client and server Windows updates available at that time. Future updates released for Windows Server 2025 and Windows 11, version 24H2 will share the same KB numbers, but will have different release note sites and links."

Regardless of whether third-party management software is to blame or Microsoft itself, plenty of admins had a tough week.

"It's not been a great week at work, too much time lost on this, and the outcome is that in some instances backups have come into play however Windows Server 2025 licensing will have to be purchased for others," the small business employee noted in their post.

The second unexpected upgrade scenario involved a banner displayed on the device's Windows Update page in Settings, offering an upgrade to the new OS.

"This message is intended for organizations that want to execute an in-place upgrade," the company said, adding that the issue had since been resolved.

Recurring issues

Microsoft quickly spotted three bugs in Server 2025 - one that meant English text was shown during installation, even if it wasn't the default language, and a second that limited the OS from running on devices with a high core count, sparking crashes, blue screens and long restarts.

Similarly, a third error popped up saying "boot device inaccessible" in iSCSI environments.

Microsoft has acknowledged all three issues, saying it is working on a resolution for each, and promising a solution in a future Windows monthly update for the core count and iSCSI bugs, and merely more information for the English text issue.

Now into its sixth generation, HPE's ProLiant MicroServer just keeps marching on. The original N36L model was introduced in 2010 and since then we've reviewed the Gen8, Gen10, and Gen10 Plus versions with the Gen10 Plus v2 model appearing around a year ago and adding support for Intel Xeon E-2300 CPUs.

The MicroServer Gen11 keeps the focus on small offices seeking a compact entry-level server for up to 15 users but also targets hybrid cloud and edge applications. It bears some similarities with the Gen10 Plus v2 but is a lot more powerful as along with a dual-core 3.7GHz Pentium Gold G7400, it's also available with a choice of two Xeon E-2400 CPUs.

Memory goes from DDR4 to faster DDR5 and the four DIMM slots in the MicroServer Gen11 allow this to be doubled to the maximum 128GB. Storage doesn't see any changes as it uses the same two groups of horizontally stacked LFF cold swap drive bays at the front.

Expansion sees big improvements as the MicroServer Gen11 offers PCIe Gen5 and Gen4 slots – the Gen10 Plus v2 only has a single Gen4 slot. Remote management also gets a boost as the system board sports HPE's classy iLO6 controller.

HPE ProLiant MicroServer Gen11 review: Clever chassis design

Sitting a Gen10 Plus next to the Gen11 model shows the new chassis is slightly larger. There's no appreciable difference in depth (4mm at most) but the Gen11 is wider by 16mm and taller by 35mm.

Its chassis cover is retained by two thumbscrews at the rear and removed by sliding it back and upwards. The front cover also has internal locking tabs on each side that can only be accessed with the lid off so if you use a padlock or Kensington lock, the drive bays are safe from wandering fingers.

HPE has put the extra room to good use and has radically improved system board access. On the Gen10 Plus, you pull a couple of cables and slide the tray out the back but accessing the Gen11 board is even easier.

With the lid and front cover out of the way, you remove two Torx screws on the top of the chassis and two more down next to the front hinges. The smart clamshell design allows the upper section to be swung over through 90 degrees to provide unfettered access to the system board – you don't even need to remove any cables.

Cooling is handled efficiently as the CPU has a large passive heatsink with two copper heat pipes routed to a vertical radiator array located in front of a 9cms diameter system fan. Noise levels are extremely low making the MicroServer Gen11 a great choice for small offices that value tranquillity.

HPE ProLiant MicroServer Gen11 review: Storage and expansion

The MicroServer Gen11 keeps the carrier-less bays from its predecessor. To install a drive, you fit mounting screws on each corner and slip it in until the tab as the front locks – sixteen Torx T15 screws are provided in the panel below.

You can fit SATA SFF drives using the optional HPE MicroServer Gen10 SFF NHP SATA Converter Kit. The quick specs guide lists this as part No. 870213-B21 and we found it on Amazon for around £12.

Intel keeps RAID duties as these are handled by its embedded VROC (virtual RAID on CPU) controller which supports RAID0, 1, 5, and 10 arrays for SATA devices. The default seeing is SATA AHCI but VROC can be easily enabled and arrays created from the server's RBSU BIOS menu.

Network connections are plentiful as the server presents four embedded Gigabit ports. There's room for more, such as multi-Gig or 10GbE, as both PCIe slots support half-length, half-height cards with the Gen5 slot the x16 variety.

HPE ProLiant MicroServer Gen11 review: Server management

The motherboard has HPE's iLO6 controller chip on-board and on our review system, we were able to load its web console using shared access from the first LOM port for in-band management on the local network. HPE's quick specs guide is a little confusing on this topic, but if you want out-of-band (OOB) remote management on a dedicated network, you'll need HPE's ML30 Gen11 Serial Port/iLO/M.2 enablement kit.

The card fits in a dedicated slot on the system board, provides a single M.2 SATA/NVMe SSD slot for a fast boot device, puts an extra Gigabit port at the back for OOB iLO6 access, and includes a cable kit for adding a serial port. It also comes with an iLO6 Essentials license which activates full OS remote control and virtual media services – the in-band option doesn't include these features and will require a license upgrade.

The iLO6 console provides all the management goodness you'll find in HPE's higher-end servers with plenty of valuable information about critical components and their status along with 3D thermal graphs showing temperature ranges throughout the chassis. The server uses the same 'dumb' external 180W PSU as the Gen10 Plus and v2 so power monitoring graphs aren't available.

The console has an option for integrating the MicroServer Gen11 with HPE's Compute Ops Management web portal – useful for larger businesses that may have them deployed to multiple remote offices. You also have HPE's Intelligent Provisioning for easy OS deployment and we used it to load Windows Server 2022 on a mirrored pair of 4TB SATA drives in 25 minutes.

HPE ProLiant MicroServer Gen11 review: Is it worth it?

The ultra-compact and powerful MicroServer Gen11 is a fine choice for small office, remote site, or edge deployments. It looks good value too, as the price for the system on review is around £890 excluding VAT. This represents a relatively modest increase over the Gen10 Plus v2 as the model with an Xeon E-2314 and 16GB of DDR4 costs around £738.

Management sees big improvements as it gets the benefit of the latest iLO6 controller although you'll need the upgrade card if you want dedicated OOB access. The MicroServer Gen11 also doubles memory capacity and the PCIe slot count, wraps it all up with a sturdily constructed chassis and its smart clamshell design makes upgrades and maintenance a piece of cake.

HPE ProLiant MicroServer Gen11 Specifications

HPE has had a sharp focus on the network edge for longer than most and it now aims to deliver even more processing power to harsh environments such as retail, manufacturing, and logistics. The ProLiant DL145 Gen11 certainly looks up to the task as this single socket (1P) server puts an AMD EPYC 8004 'Siena' CPU with up to 64 cores into a surprisingly small footprint.

This 2U system supports CPUs with a maximum 200W TDP so it supports all twelve of the 8004 family including the six 'PN' NEBS-friendly models. These CPUs are edge-optimized as they offer improved energy efficiency, lower operating costs, and reduced cooling demands.

Measuring only 381mm wide and 507mm deep with the filter bezel fitted, it's around 50% smaller than HPE's DL385 Gen11. The server is resistant to extreme shock and vibration and rated for continuous operations in temperatures between -5 and +55 degrees C.

Its high CPU core potential makes it a good choice for virtualization at the edge and it's ready for AI and ML workloads too, as it has room for up to three Nvidia L4 Tensor Core GPUs. The server supports up to 768GB of DDR5 memory and although storage capacity isn't high, it offers a choice of SFF SATA/NVMe or EDSFF E3.S SSDs.

HPE ProLiant DL145 Gen11 review: System design

For edge duties, the DL145 Gen11 employs the standard reverse system design with all expansion slots, drive bays, network ports, and PSUs accessible from the front. System cooling is handled by four standard or high-performance hot-plug fans at the rear which draw air from the front and through the filter bezel.

The plastic filter bezel slots onto the front and has apertures with dust brushes on each side for cable connections. The dust filter is easily removed with HPE recommending it be replaced every three to six months.

There's plenty going on under the lockable lid but good design means all key components are easy to access. The 48-core 2.5GHz AMD EPYC 8434P CPU in our system is topped off with an impressively chunky passive heatsink and flanked on each side by three DIMM slots which support up to 768GB using HPE's 128GB DDR5 SmartMemory modules. For power, you can choose from single and dual redundant reverse-flow 700W or 1,000W Flex-Slot PSUs.

HPE ProLiant DL145 Gen11 review: Storage and expansion

Due to their reduced size, edge servers don't have a lot of room for storage but the DL145 Gen11 manages to squeeze a good range of options in. We have the two-bay SFF drive cage which supports SATA or NVMe U.3 SSDs but you can opt for up to six EDSFF E3.S SSDs with HPE offering a four-bay cage and an additional two-bay cage that slots in above it.

You can directly attach any of the storage bays to the motherboard's SATA or MCIO connectors but AMD's chipset doesn't do integrated RAID so you'll need to source one of HPE's controller cards if you want this. HPE offers three MegaRAID-powered adapters including the MR408i-o Gen11 OCP card with all supporting SATA, SAS, and NVMe devices.

The small bay next to the drive cage is dedicated to HPE's NS204i-u Gen11 boot-optimized storage device. It's great for keeping the standard bays free for data as it presents two hot-plug M.2 NVMe SSDs and provides mirrored redundant storage for an OS or hypervisor.

The expansion bay in the center of the front panel comes as standard with one PCIe Gen5 full-height, half-length slot which was occupied in our system by an Nvidia L4 GPU. You can add two more Gen5 x16 slots with optional risers and lurking underneath the cage is the OCP 3 slot.

The server doesn't have any embedded network ports so you'll need to factor these in. Our server included a dual-port 10GbE OCP 3 card and HPE offers plenty of Gigabit, 10GbE plus 25GbE PCIe and OCP 3 cards.

HPE ProLiant DL145 Gen11 review: Remote management

The DL145 Gen11 comes with HPE's embedded iLO6 chip and a dedicated Gigabit port. The USB iLO service port is used for downloading the active health system log to a flash drive or connecting a laptop with a USB to an Ethernet adapter and directly accessing the iLO's web console.

The web interface is packed with status information on all critical components. The base price includes an iLO6 Standard license which must be upgraded to an Advanced version to enable features such as full OS remote control, virtual media services, and power management.

Security out at the edge needs to be good and the iLO6 provides plenty of valuable features such as Secure Start, 'silicon root of trust' firmware fingerprinting and SPDM (Security Protocol and Data Module) hardware component, and PCIe expansion card authentication. Physical security is in abundance too, as the chassis has intrusion detector switches, lockable panels all around, and holes for Kensington locks.

Deployment to remote sites is aided by HPE's zero-touch provisioning so businesses can deliver preconfigured servers which just need a secure network connection for swift onboarding. For ongoing remote management, the server can be integrated into HPE's GreenLake Compute Ops Management cloud service which provides a single portal for centrally managing server deployments and automating system lifecycle management and firmware compliance policies.

HPE ProLiant DL145 Gen11 review: Is it worth it?

We received an early review sample so pricing has yet to be firmed up but the base CTO chassis (SKU: P71964-B21) without CPU, memory, or storage comes in at around $2,100. Support for the entire range of EPYC 8004 CPUs means you can spec the server up to suit your budget and it supports plenty of other options.

Build quality is excellent, it's small enough to go where normal servers can't and HPE's acoustics design works well as even with a 48-core EPYC CPU and Nvidia GPU in residence, our review system was pleasingly quiet. The DL145 Gen11 is clearly capable of delivering plenty of processing power out to the edge and deployment options are extensive as HPE offers a range of desk, rack, and wall-mount kits.

HPE ProLiant DL145 Gen11 specifications

Microsoft has released the latest version of its server operating system, Windows Server 2025 — but it comes alongside a trio of bugs.

Windows Server 2025 is Microsoft's latest version of its server OS, following on from Windows Server 2022. The new version is arriving alongside System Center 2025, which Microsoft said means it's possible to "make the most” of new features immediately.

The tech giant promises better security, performance, and "cloud agility" with the latest version, but warned admins to be wary of three bugs during the installation process that include an issue that causes blue screen errors in systems with a high core count.

"Generally available today, Windows Server 2025 builds on our mission to deliver a secure and high-performance Windows Server platform tailored to meet customers’ diverse needs," wrote Ian LeGrow, corporate vice president for Azure Edge and Platform, in a blog post.

"This release will enable you to deploy apps in any environment, whether on-premises, hybrid environments, or in the cloud."

Key features in Windows Server 2025

On the security front, Microsoft has beefed up security in Active Directory, including new cryptographic support, and added new techniques for preventing a range of attacks against server message block (SMB), including hardened firewall defaults and protections against brute force, spoofing, and man in the middle attacks.

For cloud users, Microsoft highlighted a few notable features, such as hotpatching for hybrid and on-prem clouds, though this will require an additional subscription fee.

"Customers operating fully in the cloud have inherent modern security advantages like automatic software updates and back-up and recovery," LeGrow wrote.

"Now we’re bringing some of those capabilities to Windows Server 2025 for on-premises customers with a new hotpatching subscription service, enabled by Azure Arc. With hotpatching, customers will experience fewer reboots and minimal disruption to operations."

Alongside that, Microsoft has integrated Azure Arc for easier onboarding and operational flexibility, unified network policy management, and software-defined network multisite features for easier migration and management.

Microsoft of course can't fail to mention AI. LeGrow said Windows Server 2025 was designed for demanding workloads including AI and machine learning.

"Windows Server 2025 delivers up to 60% more storage IOPs performance compared to Windows Server 2022 on identical systems," he said. "Windows Server 2025 introduces massive performance and scalability improvements that come from Azure."

Alongside the new features, Microsoft is also killing off a few. That includes WordPad, Windows PowerShell 2.0 engine, and the SMTP service.

Watch out for the bugs

Microsoft has warned about three particular issues that may crop up during installation. The tech giant confirmed it is working on addressing all three bugs, which have varying levels of seriousness.

The first bug might show English text during installation, regardless of the language selected. As Microsoft points out, that won't affect English-language users, but "will be noticeable if a language other than English is selected for installation."

However, this error only happens if using a CD or USB drive to install Windows Server 2025.

The second means Windows Server 2025 might not run properly on devices with a high core count. Microsoft said it appears to be happening on servers with more than 256 cores, and is causing the installation or upgrade to fail or stop, long restarts of three hours or more, and a blue screen error.

The issue doesn't seem to happen on all systems with high core counts, but if it does, limit the number of cores below 256 for installation.

Finally, the third issue includes an error saying "boot device inaccessible" may be displayed in iSCSI environments during startup after installation of Windows Server 2025.

Key details

Windows Server 2025 is available immediately, alongside System Centre 2025. As a long-term release, it will have extended support until 2034, and mainstream support until 2029.

It is possible to upgrade directly to Windows Server 2025 from as far back as Windows Server 2012 R2. Windows Server 2022 will be supported until October 2026.

Windows Server 2025 will work with most 64-bit processors, and Microsoft has released a full list of certified processors.

Small businesses still storing and sharing data across separate office PCs need to invest in a server if they want to become more efficient and productive. Servers are powerful, purpose-built systems designed to provide a wide range of services including network storage, file sharing, application hosting, messaging, remote access, and data security.

Despite the allure of cloud-hosted services, on-premises servers offer significant benefits to small businesses. Cloud outages are becoming ever more frequent but with the appropriate core infrastructure systems in place, in-house servers can deliver 24/7 availability and put small businesses totally in charge of security, services, and business applications.

There are many other compelling arguments as you own the servers so you can customize and upgrade them whenever you need to and securely store data and share it across your workforce. Budgeting and cost management are other key considerations as although your initial investment will be higher, you won't be faced with monthly cloud subscription fees so ongoing operational expenses will be significantly lower.

Entry-level servers are affordable for small businesses and a solid long-term investment that can grow with you for many years. Here, we show how to set up a server for small businesses and the features you should be looking for.

Budget responsibility

Budgets are tight in these trying times but a common mistake small businesses make when acquiring servers is spending too little. Cutting costs at this stage is a false economy as a poorly specified and underpowered server will struggle to cope with demand as a business grows and will most likely need replacing in the near future.

A key feature of a server is hardware redundancy so it will keep running in the event of a component failure. Purpose-built servers come with business-grade storage devices, support multiple RAID arrays as standard, have sophisticated cooling systems comprising multiple redundant fans, and often include (or support) dual redundant power supplies – features you won't find in a basic PC.

Choose well and your servers can last you a surprisingly long time. One of our lab test servers is a trusty Dell PowerEdge T640 tower server which we specced up with the future in mind, and even after six years it's still functioning perfectly and providing a sterling service as a Windows Hyper-V virtualization platform.

Small businesses should also curb their enthusiasm and avoid spending too much. It's easy to get carried away with the latest and greatest in server technology but it all comes at a price and if the server spends most of the time twiddling its thumbs then it is a wasted investment.

Know your workloads

There is a wide range of small business servers on today's market but before jumping in, it's important to understand what you want the server to do - not just now, but well into the future. Ask yourself what workloads you plan to centralize, the type of resources they will require, your long-term business goals, and what headroom you want for future expansion.

If the server will function as a central storage vault, provide file-sharing services, and maybe host databases, then you'll need plenty of internal capacity. Server storage can get swallowed up very quickly so choose servers that have plenty of drive bays and factor in spare capacity as it's easier and a lot cheaper to expand an existing RAID array with extra disks than it is to replace the entire server.

Location is also very important as this will determine the server form factor. Small businesses that only need a single server will find tower models a good choice as they don't need any additional infrastructure such as rack cabinets and can be placed wherever you want.

Low-profile 1U (44.5mm high) rack servers offer more flexible options as multiple systems can be placed in rack cabinets that require less floor space. You can scale up on-demand by adding more servers to the cabinet and if you need more expansion potential and higher storage densities consider 2U high rack servers.

Processor decisions

For most small business workloads, a tower or rack server with a single processor socket (1P) represents the best value and there's plenty of CPU models to choose from. Intel has traditionally had the strongest presence in the entry-level server market with its Xeon E CPUs.

The current Xeon E-2400 family consists of eight CPUs with base speeds ranging from 2.6GHz up to 3.5GHz and core counts going from four to eight with all bar the entry Xeon E-2414 supporting Intel's HT (hyperthreading) technology. Maximum memory for all Xeon E-2400 CPUs is 128GB of fast DDR5 which will be perfectly adequate for most small businesses

For more demanding workloads, you may want to consider AMD's new EPYC 4004 CPUs. Only recently launched, this group of eight CPUs are pitched directly at the cost-conscious small business server market.

The family also comprises eight models but offers much higher clock speeds ranging from 3.7GHz up to 4.5GHz and all support a more generous 192GB of DDR5 memory. The two higher-end models also fill a gap left by the Xeon E-2400 CPUs as they offer 12/24 and 16/32 cores/threads.

Protect and serve

A server is a critical part of business operations so it needs to be protected. RAID will cover you against disk failures but also consider dual power supply units (PSUs) as if one fails, the second will keep the server running.

Invest in a UPS (uninterruptible power supply) as this will keep your server safe from harm during a power spike or short-term power outage. There are plenty of affordable models suitable for small businesses with the distributor Broadbandbuyer offering a wide range of products from APC, Eaton, and CyberPower.

Remember too that RAID and power redundancy will protect your hardware but aren't data backup solutions. We strongly recommend getting a proper server backup strategy running right from the outset - there is a wealth of small business backup products available with many combining on-premises with cloud backup to ensure you have essential off-site copies available.

When sourcing a small business server don't skimp on the warranty and maintenance contracts either. Blue chip server vendors such as Dell and HPE usually offer three-year warranties as standard and this includes a next business day (NBD) on-site support service which can be upgraded to a four-hour response for critical systems.

And when your server is up and running, you'll want to keep a close eye on its health so monitoring and management features are equally important. Even small business servers have specialized controllers on their motherboards that provide a remote web console so you can see the status of critical components in real-time with the best ones providing timely email alerts if anything untoward is detected.

Selecting your first server needn't be challenging and here at ITPro, we've reviewed most of the current entry-level tower and rack models from big-name vendors including Dell and HPE. These are ideal for small businesses as they can help centralize multiple IT functions on a single platform, reduce acquisition costs, boost productivity, and deliver data availability and security.

At the board level, sustainability is becoming a serious talking point. With generative AI driving demand for data centers and in turn pushing up emissions, leaders have to think about their company’s carbon footprint.

Ben Holme, associate CIO for infrastructure at the Univeristy of St Andrews, is one such leader. His role demands he take care of the university’s cloud and networking systems, as well as its servers, storage, and data centers.

Holme tells ITPro that just as everyone plays a role when it comes to cybersecurity, every department also has a responsibility to help their organization hit its sustainability targets.

“Sustainability is really, really important to us,” Holme says. “We are teaching the leaders of tomorrow, so it's important that we entrench this in students,” he adds, explaining that the institution tries to embed it into everything it does.

This was ultimately the improvement the University of St Andrews saw when it made the move to Pure Storage’s flash solutions, despite it not being the initial trigger for making the switch.

“When I started in the role five years ago, we had Dell storage solutions,” Holme says. “We had a flash-based solution for our block storage platform, and we had … disk-based solution for file.”

Around five years ago the previous disk-based solution was updated to a disk-based solution that operated with a very small amount of flash acceleration in the front end. Holme was in charge of reconsidering the university’s approach to storage, driven largely by spiraling bills.

“At the time, cost was a big factor, because extending the incumbent platform in both support and capacity was actually quite expensive,” he tells ITPro. “So it made us stop and think if that's what we wanted to continue doing.”

After conducting an analysis and seeing what else was available, Holme and his department opted for Pure Storage’s solution: a pair of FlashArray// X20s, with an eventual upgrade to X70s.

While the price of storage was the spark for change, Pure’s message on sustainability helped to reel St Andrews in as a client. “There was a very clear message around Evergreen and it felt a genuine message for me, it wasn't just marketing,” he says.

He adds that Pure was the most efficient option on the market for St Andrews, as the other two options in contention required either one and a half times or two times the energy to run without any added peripherals such as network switches or virtual machines (VMs).

Over the course of ten years, the University of St Andrews is predicted to save around 150,000 kilowatt hours (kWh) of electricity through Pure Storage’s solutions. That’s enough to power 42 homes for a full year per Ovo figures, or over 1,000 desktop computers for the full decade per Energy Sage.

Holme notes he was also drawn to the potential escape from step migrations. If using the Pure platform worked out for the university, it could avoid doing any step migrations in the future.

A seamless transition

Holme struggles to recall any bumps in the road when it comes to working with Pure.

“The Array//Xs have basically sat there and done their thing,” he tells ITPro. “The only excitement we had on them recently was we had to move one of them,” Holme says. Even this was a simple process, as the storage system was able to be powered down by a simple pulling of the plug.

This, Holme says, came as a shock. After 10-20 years working with storage systems, having always made sure to shut storage down clean, and drain the caches, just unplugging the unit was a very different experience.

“They said unplug it. We unplugged it. They said, ‘If you put it back within – maybe it's two hours, I forget now – all of the ActiveSync state will stay and it will just carry on as if nothing had ever happened. And it did.”

Looking to the future, Holme says St Andrews is paying close attention to new Pure offerings, including active cluster for file and enhanced file services through Pure Storage Fusion.

“The base product gives us what we need, I think that's the first thing to say,” Holme adds. “I have the confidence that we have the performance, the current capacity, but also empty slots to increase that capacity, certainly for the next three to five-year window.”

A bare metal server is one where you have effective ownership of the entire computer, and can control exactly what hardware is installed and what software runs on it. The name reflects the fact that you have access to the "bare metal" of the device.

In the past, when companies operated their own server rooms, bare metal configurations were the norm. The past few decades however have seen most businesses move to cloud-hosted virtual servers, and bare metal is now considered a specialist requirement

What's the difference between bare metal and a virtual server?

A virtual server operates entirely inside a virtual machine. Its access to the host hardware is managed and mediated by a hypervisor application. When you lease a cloud-hosted virtual server, it will probably be one of multiple virtual machines running on a single physical computer – but you will only have access to the specific virtual server that's running your software, which in turn will only have access to computing resources that the hypervisor makes available.

In a bare metal setup, by contrast, your server software runs directly on the hardware and has full access to all resources. You may choose to install a virtualization host and run virtual machines inside the server environment, but the whole computer is still ultimately managed by you.

Is bare metal the same as a dedicated server?

For an extra fee, some companies offer dedicated servers (also known as single-tenant servers), which come with a guarantee that no virtual machines other than your own will be running on the provided hardware.

This allows you to avoid "noisy neighbor" syndrome – where your server performance is affected by other virtual machines running on the same physical hardware and tying up the available CPU, RAM, or storage.

However, these dedicated server offerings may still involve accessing a virtual environment, configured to the hosting provider's preferred parameters and running on a standard hardware platform. With a bare metal server, you're completely free to configure the hardware and software as you please.

What are the advantages of configuring your own server hardware?

While virtual servers can be highly flexible, bare metal gives you more options to precisely tune the hardware to suit your specific workload, such as installing expansion cards to accelerate particular operations.

The ability to customize the hardware is also useful for developers who need to test their code's compatibility with specific devices and peripherals – or those working with low-level drivers, which aren't normally accessible from within a virtual machine.

Why might you want a custom software specification?

Depending on your workload, you may see performance gains from running your operating system and software directly on the hardware, rather than inside a hypervisor. With a bare metal configuration, there's also less chance of a security breach arising from bugs or vulnerabilities in the hypervisor, or in the hosting provider's orchestration software.

At the same time, it's important to remember that you alone are responsible for the security and the safety of your data, and that applies especially with bare metal servers. A hosting provider's standard backup and data replication services may not be available on non-standard platforms, and they might have a very limited ability to detect and deal with any security issues that may arise.

How do you get access to a bare metal server?

It's still entirely possible to set up a bare metal server on your own premises and access it over your own network. This gives you direct access to the hardware and storage devices, so you can very easily configure and adjust the configuration as needed.

Alternatively, many hosting providers now offer bare metal servers inside their own data centers. Even larger platforms can provide such resources on request, within certain parameters: for example, both Amazon EC2 and Microsoft Azure offer bare metal servers based on Intel Xeon Scalable processors with between two and 24 sockets.

Using a hosted bare metal server means you keep some of the advantages of cloud infrastructure, such as the ability to scale your resources up or down as required, without needing to invest in expensive processor and hardware components. You can also benefit from very fast connections between the bare-metal server and virtual machines hosted on the same network.

Other benefits include protection against power outages, hardware failures, and other material risks. Depending on your configuration, you may be able to set up backups and data redundancy too, for a high degree of confidence that even if the worst should happen your data will be protected.

At first glance, Broadberry's CyberServe EPYC EP1-226T already looks to be a well-specified and versatile server but closer inspection reveals some big surprises in the storage department. This 2U rack server supports NVMe SSDs but avoids the performance pitfalls of hardware RAID controllers by showcasing the SupremeRAID SR-1010 RAID solution from GRAID Technology.

Hardware RAID cards have a problem with the latest high-speed NVMe SSDs as their PCIe interfaces can present a performance bottleneck when handling even a small pool of these devices. Traditional software-managed RAID is a cheaper alternative but also has many limitations, not least with its voracious appetite for CPU resources.

GRAID Technology neatly overcomes these issues by handing all the hard work over to a GPU with the SupremeRAID SR-1010 based around a low-cost but powerful Nvidia RTX A2000 PCIe 4 double-width card. Essentially, it is a software-defined solution that offloads all RAID operations to the GPU and uses out-of-path technology to significantly reduce CPU load.

The SupremeRAID SR-1010 has a lot to offer as it supports up to 32 direct-attached NVMe SSDs and can manage RAID0, 1, 10, 5, and 6 arrays. It can be deployed on a wide range of Linux and Windows OSes and upcoming software updates aim to add features such as compression, encryption, and thin provisioning.

Broadberry CyberServe EPYC EP1-226T review: What's inside?

The CyberServe is built on a solid foundation which comes courtesy of Asus' latest RS520A-E12-RS24U server system. This 1P server is powered by a 2.9GHz 24-core AMD EPYC 9254 CPU and partnered by 128GB of DDR5 memory which can be pushed to a maximum of 6TB.

Cracking the lid reveals a remarkably tidy interior mainly since the SupremeRAID card doesn't require any extra cabling. The drive backplane is cabled directly to the motherboard's embedded MCIO (mini cool edge IO) ports with the SupremeRAID card sitting comfortably on its own in the left rear x16 PCIe expansion riser.

In this configuration, the server supports a maximum of 16 NVMe SSDs. The EPYC CPU presents 128 lanes with 4 taken for each SSD, 16 going to the SupremeRAID card and the rest dished out to the other PCIe Gen 5 slots, the OCP 3 Gen 5 slot, and associated devices such as the motherboard's dual M.2 NVMe/SATA SSD slots and embedded dual Gigabit LOM.

The remaining eight drive bays can be gainfully employed by fitting standard SAS/SATA SFF devices. Your choice of OS is covered too, as Broadberry includes a rear 2-bay SFF cage fitted with 480GB Samsung PM893 SATA SSDs and managed by a Broadcom 9461-8i RAID card.

Other features of note are remote management as the server comes with an Asus ASBM11-iKVM controller chip and dedicated Gigabit port. Its web interface presents everything you need to know about server health and multiple servers can be centrally managed with the Asus Control Center web application.

Broadberry CyberServe EPYC EP1-226T review: GRAID configuration

Broadberry supplied our system with Windows Server 2022 pre-loaded on the mirrored SATA SSDs and primed with the Nvidia and SupremeRAID Windows drivers. The card comes preset to Nvidia's compute mode and Broadberry can preconfigure the NVMe SSDs to your requirements or you can do it yourself from the command prompt.

It's easy enough as you first run one command to declare the SSDs to the SupremeRAID card as physical drives (PDs), after which they are un-bonded from the OS. Now you can create a drive group (DG), specify which managed SSDs are to be included and the desired RAID array type – the card currently supports up to four DGs in Windows environments.

Lastly, you create virtual drives (VDs) which then appear as storage devices to the OS with up to eight VDs per DG supported. If required, a consistency check service can be run manually on the VDs or scheduled at regular intervals and set to automatically correct any errors.

Another advantage is the SupremeRAID card doesn't require additional cache memory and bulky battery backup units. The software needs to be licensed and Broadberry advised us the price of this system includes a perpetual license.

Broadberry CyberServe EPYC EP1-226T review: Stonking performance

Broadberry included twelve 3.8TB Micron 7450 PRO U.3 NVMe SSDs which only took a minute to declare to the SupremeRAID card. We created a RAID6 DG using them all and configured a single VD which appeared to the OS as a new 35TB drive.

For performance testing, we started with Iometer configured for 32 disk workers, a queue depth of 32, and a 1MB transfer request size. Prepare to be amazed as the VD returned a remarkable 58GB/sec and 21GB/sec for sequential reads and writes.

Moving to random operations saw no letup in speed with Iometer reporting 56.2GB/sec for reads and 20.6GB/sec for writes. Swapping Iometer over to a 4K request size delivered equally impressive I/O rates with sequential reads and writes delivering 1.11 million IOPS and 591,100 IOPS while random reads and writes both returned around 1.1 million IOPS.

We also noted during testing that the SupremeRAID card was doing its job as server CPU utilization never went above 33%. Performance on our system was extremely good but it'll go even faster under Linux with GRAID Technology claiming maximum speeds for a 24 NVMe SSD RAID6 array of 260GB/sec for sequential reads and 28million IOPS for random reads.

Virtualization platform support doesn't include VMware but Broadcom's new licensing schemes are forcing many businesses to look elsewhere anyway. They'll be pleased to know that all the SupremeRAID products support Microsoft Hyper-V, Linux KVM, OpenVZ, and the increasingly popular Proxmox.

Broadberry CyberServe EPYC EP1-226T review: Is it worth it?

With the system on review costing £15,906 ex VAT, the CyberServe EPYC EP1-226T packs in a heck of a lot of hardware for your money. It works for us as a powerful general-purpose storage server but the SupremeRAID GPU card makes its credentials even more appealing.

GRAID Technology is a well-established company as it was founded in 2020 and along with the SR-1010 for 2U rack servers and SR-1000 for 1U servers, it has launched the entry-level SR-1001 which targets edge and tower servers with up to 8 NVMe SSDs. Our lab tests show the SupremeRAID SR-1010 clearly delivers on its performance promises and makes the CyberServe ideal if you want to get the best from your investment in NVMe SSD RAID arrays.

Broadberry CyberServe EPYC EP1-226T specifications

You might not think much about Domain Name Servers – or DNS for short – but you probably access them every day. These are the servers that work behind the scenes to translate familiar domain names (such as itpro.co.uk) into the numeric IP addresses used by the internet's backbone servers.

Normally these DNS lookups happen automatically, using servers operated by your ISP – but if you wish you can change your settings to use any available servers.

Why should I change my DNS?

Your ISP's name servers can normally provide a very fast response because they're connected to the same core network as your client devices. However, if the server is overloaded, or if its network connection is congested, you could still experience a slow service. If you're noticing that online sites and services are slow to load when you first enter their names then switching to a different name server could help. (Subsequent connections should be fast regardless of DNS performance because once your computer receives the addressing information it will cache it for future use.)

Another reason you might choose to change your DNS settings is security. It's theoretically possible for cybercriminals to hack into a poorly managed name server and change the addressing rules so that legitimate domain names are redirected to malicious sites. If you suspect your DNS might be vulnerable to such an attack, you can protect yourself by moving to a trusted server.

Some DNS operators also maintain lists of sites that are known to be harmful, and block attempts to connect to them. Switching to these servers provides you with instant protection against phishing scams and other online dangers.

Which DNS should I use?

Online devices usually have two DNS addresses configured, one primary and one secondary. If the primary DNS doesn't respond to a query in a reasonable amount of time, the device will automatically try the secondary one instead. If you're finding your regular DNS is slow, you can try swapping these around, so you're contacting the other server first.

Alternatively, you can replace your ISP's servers with a free public DNS. One of the simplest options is Google Public DNS, which promises a fast, robust service: to try it out, you simply need to configure 8.8.8.8 as your primary DNS address, and 8.8.4.4 as your secondary server (see below for details of how to do this).

Cloudflare is another company offering a free name service similar to Google's, using 1.1.1.1 and 1.0.0.1 as its primary and secondary servers. It also offers services with automatic malware blocking on 1.1.1.2 / 1.0.0.2, and adult content filtering on 1.1.1.3 / 1.0.0.3.

One of the most versatile DNS options is OpenDNS, operated by network specialist Cisco. It provides free DNS at IP addresses 208.67.222.222 and 208.67.220.220, and allows you to configure optional blocks on phishing, adult sites, and other content. For business networks, DNS-level security is also part of the Cisco Umbrella security platform.

How to change DNS settings in Windows 10

Network clients are usually configured to send DNS queries to the local internet router, which then takes care of forwarding them to the correct server. This approach means you can move your whole network onto a different domain name service by simply updating the DNS addresses in your router's management interface.

If you prefer though you can configure individual computers to use specific name servers. In Windows 10 this is easily done using either the Command Prompt or the Control Panel.

To do it via the Command Prompt, start by opening an elevated shell: a quick way to do this is by opening the Start menu, typing "cmd", then selecting Run as administrator, and accepting the security prompt that appears.

Once the Command Prompt window opens, enter the following command to see a list of network interfaces:

netsh interface show interface

Make a note of the name of the network whose DNS settings you want to change (in our case it's "WiFi"). You can now configure this network to use whatever primary and secondary DNS servers by entering the two commands below – inserting the name of your own network interface between the quotation marks, and replacing x.x.x.x and y.y.y.y with the IP addresses of the primary and secondary DNS you want to use:

netsh interface ipv4 set dns name="Network Name" static x.x.x.x primary

netsh interface ipv4 add dns name="Network Name" y.y.y.y index=2

The change should instantly take effect. You can check your active DNS settings by entering:

netsh interface ipv4 showdnsservers

How to change DNS settings using the Control Panel

If you prefer to use a graphical interface to change your DNS settings, you can jump straight to the relevant page of the Control Panel by hitting the Start button and searching for "View Network Connections". Selecting this will bring up a window showing all configured networks, which might include Bluetooth and virtual network adaptors as well as Wi-Fi and Ethernet connections.

Now right-click on the network you want to configure and select Properties to open a window showing the various drivers and protocols used by that connection. Double-click on "Internet Protocol Version 4 (TCP/IPv4)" (or single-click it and hit the Properties button below) to open a new window showing your IPv4 address settings.

In the lower half of this window, click on "Use the following DNS server addresses", then enter your desired primary and secondary addresses into the relevant fields. Hit OK to close the window, then OK again on the main connection properties dialogue.

Your DNS will now be immediately updated, and you can close the Control Panel. If you want to go back to router-based DNS, just come back to this page and select "Obtain an IP address automatically".

How to change DNS settings in Windows 11

If you're fed up of controls and features changing between different versions of Windows, you'll be pleased to learn that the process for changing DNS settings in Windows 11 is precisely the same as in Windows 10.

That applies whether you're using the Command Prompt or the Control Panel – so you can follow our instructions above to set up your Windows 11 system in whatever way you like.

You will likely be familiar with the concept of the server, a set of computers that underpin the vast majority of the digital systems within your organization. Servers provide the resources, data, and services used by all of the other computers on the network, and as such are a vital part of any IT infrastructure.

Designed to be highly reliable, they can serve a wide range of purposes within a business from hosting websites, storing files, managing databases and email systems, and many more. However, there are a variety of different ways in which organizations can choose to set up their server ecosystem, with a varied selection of server types to pick from. These include dedicated servers, shared servers, virtual private servers (VPNs), cloud servers, and more.

Dedicated servers are physical servers dedicated to a single business customer, hence the name. Synonymous with bare-metal servers, dedicated hosting services, or managed hosting services, all of which refer to the fact that the software and hardware encompassed by the server is managed by a company for a single client.

As a result, the client has total control over the server's resources, including its CPU, RAM, or GPU. Organizations can optimize the hardware of the server and allocate these resources according to their needs, a very useful feature.

The benefits of using a dedicated server

The benefits of running a dedicated server primarily stem from the fact that you have complete control over the devices, its applications, and operating systems. Here are some examples of how you can improve your IT estate with the added control dedicated servers offer.

Performance

Firstly, having sole access to a server will help you unlock better performance, since the server's resources are not being shared with other clients who could be hogging some of its computing power and slowing down your applications.

In addition, dedicated servers can be optimized according to the specific workload you plan on using it for. All of this can help your organization achieve faster load times and an improved user experience overall.

Flexibility and Scalability

Flexibility is a large part of the value proposition that dedicated servers make. They can be used to deploy a wide range of applications, and as the system is totally customizable, firms can alter the hardware and software as their needs change in the future.

For the same reason, dedicated servers scale very well as your organization expands. Firms can adapt to changing business requirements and increase their server's memory, storage, and processing power whenever they need the extra headroom.

Security

Customizability also brings security benefits too. With total control over the server configuration, businesses can add whatever protections they want to their system. These could include firewalls, encryption, intrusion detection systems (IDS),

Since the server isn't being shared with anyone else, there is also a reduced risk of being breached via a third party, further improving the overall resilience of your IT infrastructure.

Reliability

Dedicated servers also tend to be more reliable than their shared counterparts, largely because you are less likely to be affected by other users' activities. This means by opting for a dedicated server firms can enjoy extended availability and avoid costly downtime.

What are dedicated servers used for?

As mentioned above, dedicated servers are incredibly flexible and can be used for myriad IT tasks you may have. Here are some of the most common use cases for dedicated servers.

Data backups

Dedicated servers can be used to simply store backups of sensitive data, perhaps if your business operates in a sector with particular regulatory obligations or privacy concerns. For example, many healthcare and financial institutions use dedicated servers protected with added security layers to store medical records or financial data.

Virtualization

Virtualization is a very handy technique for businesses to get more bang for their buck when it comes to hardware, and firms can opt to use the tool to get more with their dedicated servers. Virtual machines (VMs) can help an organization achieve more efficient use of their resources, as well as the ability

Organizations could run a series of virtual machines (VMs) using the hardware of a dedicated server, for instance, allowing them to run different operating systems and applications on each.

Development and testing

Firms with development teams can use dedicated servers as an isolated environment in which to test their applications, without the risk of impacting production environments. This isolation also ensures that the results from the testing are accurate and not influenced by any external factors.

Devs can also alter the specific server, changing the hardware or operating system to test how their application performs in different environments, leading to smoother deployments in the long run.

E-commerce platforms

Many companies use dedicated servers to process the large volume of transactions on their e-commerce platforms. The added availability these systems can offer is paramount for businesses that need to keep their systems up through periods of very high throughput, handling vast numbers of concurrent users, product databases, as well as transactions themselves.

Small businesses seeking an affordable entry-level server have a wide range to choose from but most tend to be black, blocky, and, dare we say, a bit boring. Decked out in eye-catching silver, Dell's PowerEdge T160 breaks the mould as it packs an impressive hardware package into a remarkably compact and stylish chassis.

Commendably, Dell aims to reduce the server's environmental impact by avoiding metallic paint and leaving the steel chassis unfinished. It has a zinc coating for protection and although Dell doesn't give a figure, uses some recycled steel.

Stepping up as the replacement for the PowerEdge T150, it is 42% smaller and supports the latest Intel Xeon E-2400 CPUs and up to 128GB of DDR5 memory. Storage sees significant improvements as whereas the T150 supports up to four drives, the T160 has room for a total of seven devices – more on that later.

Another reason for the T160's industrial-strength construction is Dell is aiming it at 'near edge' applications such as warehouses and retail offices and offers an optional bezel kit with a dust filter. It's also protected from knocks and bumps as each corner has a plastic cover that doubles up as support if you want to stack multiple units on their side.

Dell PowerEdge T160 review: What's inside?

You'll get a good idea of build quality when you pick the T160 up. It may be significantly smaller than the T150 but a fully populated T160 weighs almost the same at around 11.6kgs.

The metal side panel is retained with two thumbscrews and also incorporates a Kensington lock tab at the back for extra physical security. Underneath you'll find a single plastic shroud that completely encloses the motherboard and upper front drive bays for improved air flow.

With the shroud removed, you'll find a busy interior although cabling is kept neat and tidy with all key components easily accessible. The CPU has a large passive heatsink and cooling is handled by a single 9cms standard fan behind it. 

Four DIMM slots sit below the CPU allowing the full 128GB of memory to be realised. The server has an integral LOM (LAN on motherboard) presenting dual Gigabit ports and there's room to expand as the motherboard has two PCIe Gen4 slots at the top with both supporting half-length, half-height cards.

Dell PowerEdge T160 review: Storage surprises 

Main storage is handled by a large internal cage at the front with room for three cabled LFF hard disks. It's a cinch to access as after removing one thumbscrew, the cage can be flipped down and a smart design feature is the tension spring at its base which stops it falling down and hitting the desk.

Up above are two more bays with plastic carriers for SFF drives such as SATA SSDs. Base systems will have the drive cages cabled to the server's embedded PERC S160 controller which offers software-managed stripes, mirrors, RAID10, and 5 arrays for SATA drives.

If you want more, the T160 supports Dell's PERC H355 card which offers stripes or mirrors for SAS and SATA drives, or the higher-end H755 which adds RAID5, 6, 50, and 60 arrays to the storage equation. The non-RAID HBA355i adapter provides simple pass-through services for storage devices.

That's a total of five drives so where are the other two? Amazingly, there's room below the PCIe slots for Dell's optional BOSS-N1 card which presents two M.2 NVMe SSDs in hot-plug carriers accessible at the rear. 

The BOSS-N1 is ideal for running an OS or hypervisor on a fault-tolerant mirrored array and it keeps all the main bays free for data storage. Bear in mind you can't prevent external access to the NVMe SSD carriers so if the server is out at the edge, you may want to keep it in a secure cabinet.

If you add the BOSS-N1, a PCIe card, or SFF drives in the upper cage, you'll need the extra high-performance fan unit. Fitted behind the upper drive bay, it does increase noise levels but these shouldn't be noticeable in a busy office and definitely not in a factory or retail environment.

Dell PowerEdge T360 review: Abundant management features

Good remote management is normally a feature you don't often see in a low-cost entry-level server but the T160 bucks the trend as it comes with Dell's iDRAC9 embedded controller and dedicated Gigabit port. This delivers the same features as found in Dell's higher-end servers with its web console presenting a wealth of information about system and component status along with direct access to the server's BIOS settings.

The T160 includes an iDRAC9 Basic license in the price which enables web access, hardware monitoring, and remote power controls. It's worth upgrading to an iDRAC9 Enterprise license as this enables virtual media services and full OS remote control.

A downside of the PERC S160 controller becomes apparent when you select the web console's storage tab. The S160 software controller won't be listed as the iDRAC9 only supports inventory and management of Dell's HBA and PERC RAID add-in cards plus the BOSS-N1.

Dell's embedded LifeCycle Controller makes light work of OS installation. Selected from the BIOS boot screen, we used it to install Windows Server 2022 from a mapped ISO file in only 30 minutes and it ensured all the correct drivers were loaded.

Dell PowerEdge T160 review: Is it worth it?

Absolutely. With prices starting at a shade under a grand, the PowerEdge T160 is a top-value tower server perfectly suited to small businesses and those that want to push some compute power nearer to the network edge.

Dell's design is exemplary as size for size, there's no other entry-level server that can match its hardware specification. Support for Intel's Xeon E-2400 CPUs and DDR5 memory provide plenty of power, it offers a versatile range of storage options, the embedded iDRAC9 delivers classy remote management and its slinky silver surfaces ensure it stands out from the crowd.

Dell PowerEdge T160 specifications 

HPE's ProLiant DL365 Gen11 is a rack-dense server solution aimed at enterprises that want maximum processing power in minimum space. With support for AMD's 128-core Gen4 EPYC CPUs, this dual-socket (2P) system can put an incredible 256 physical cores in only 1U of rack height.

Targeting a diverse range of workloads such as general-purpose virtualization, VDI, EAD, and CAD, the DL365 Gen11 offers a lot more than a high processing density. It can present a whopping 6TB of DDR5 memory and has room for up to two single-width or double-width GPUs.

Storage features are undiminished by its compact dimensions as there's room up front for up to ten SFF drive bays and the server supports SATA, SAS4, and U.3 NVMe devices. Even better, HPE offers a CTO (configure to order) model with a new front panel design that can handle up to twenty EDSFF E3.S NVMe SSD storage devices.

HPE ProLiant DL365 Gen11 review: Platform security and management

HPE sharply focuses on platform security as it leverages AMD's Secure Processor technology. Using HPE's exclusive iLO6 ASIC, the server validates firmware using its Silicon Root of Trust which creates an immutable fingerprint that the Secure Processor must match before the server is allowed to boot.

Consequently, the server will only boot if it detects hardware-validated HPE firmware. It uses a cryptographically signed signature so if the firmware is modified or compromised in any way, it will cause a fingerprint mismatch and the server can automatically recover to a known and trusted firmware version.

General management features are abundant with the iLO6 presenting a smart web console that's packed with valuable information about the server's status. Stepping up to an iLO Advanced license enables features such as virtual media services, full OS remote control, and power metering.

HPE's OneView monitors and manages all your HPE servers, storage, and network infrastructure devices while the free iLO Amplifier Pack presents a dedicated web console for discovery and inventory of up to 10,000 Gen8, Gen9, Gen10, and Gen11 servers.

And then there's HPE's GreenLake Compute Ops Management (COM) cloud service which is ideal for organizations with geographically distributed on-premises, cloud, and edge compute resources. All Gen10 and Gen11 servers connect directly to the COM portal which can be used to centrally manage server deployments and automate system lifecycle management and firmware compliance policies.

HPE ProLiant DL365 Gen11 review: Features and design

The DL365 Gen11 is good value with prices for the base models starting at around £5K. Designed to get you out of the starting blocks, the entry-level P59707-421 model sports a 16-core, 3GHz AMD EPYC 9124 CPU, 32GB of DDR5 memory, eight SSF bays (no storage included), a quad-port Gigabit OCP 3 card, an MR408i-o OCP 3 RAID card, and a meaty 1000W Titanium power supply.

Our review system was supplied as a P53933-B21 CTO chassis furnished with a pair of the same CPUs as the base model along with 128GB of DDR5 Smart Memory and dual 800W Platinum PSUs. You can easily tailor the system to your budget as HPE offers a wide choice of Gen4 EPYC CPUs ranging from the CPUs we have up to the mighty 128-core EPYC 9754. 

With 24 DIMM slots in play, you can expand memory up to 6TB but deep pockets are required as 256GB octal-rank RDIMM modules cost around £20K a pop and you can't mix them with non-3DS memory. Upgrades are easy as the server presents a very tidy interior with the CPUs topped off with large passive heatsinks that extend radiator bars in front of their attendant banks of DIMM slots.

Cooling is handled by a bank of up to seven hot plug fans behind the drive backplane. Our CPUs have 200W TDPs so we can use the standard fans but those above 240W will require the performance versions. 

HPE ProLiant DL365 Gen11 review: Storage and expansion

Our system had eight SFF drive bays and HPE filled them all with 1.92TB SATA SSDs which appeared as individual drives because the backplane was cabled directly to the motherboard. Unlike the DL365 Gen10 Plus model, you don't get HPE's embedded SR100i Gen10 controller chip so you'll need to specify a RAID card with HPE offering its SR (SmartRAID) and MR (MegaRAID) storage controllers with PCIe and OCP versions available.

A CTO model for GPUs is available with the server supporting Nvidia's A16 double-width and L4 single-width cards. There's another CTO option for the twenty EDSFF bays where you'll also need to specify dual CPUs plus performance fans and you'll be restricted to twelve EDSFF bays for CPUs with a TDP equal to or greater than 320W.

You can maximize front bay data storage by selecting HPE's NS204i-u Gen11 boot-optimized storage device. Accessed from the rear, it doesn't require a PCIe slot and comes with two hot-plug 480GB M.2 NVMe SSDs preconfigured in a mirrored array which we used to load Windows Server 2022.

There's plenty of room for further expansion as the server offers two PCIe Gen5 risers with a pair of OCP 3 slots underneath them. A lack of embedded network ports means you'll need to use them with HPE offering a big choice of Gigabit, 10GbE, 10/25GbE, 100GbE, and 200GbE PCIe and OCP mezzanine cards.

HPE ProLiant DL365 Gen11 review: Is it worth it?

With AMD's core-heavy Gen4 EPYC CPUs in charge, the ProLiant DL385 Gen11 is clearly capable of handling a wide range of demanding enterprise workloads. It puts a lot of processing power in small rack spaces and teams this up with tight platform security, flexible storage configurations, plenty of expansion potential, and a choice selection of remote management and monitoring tools.

HPE ProLiant DL365 Gen11 specifications

Dell's PowerEdge T360 has its sights set on growing SMBs and remote offices looking to upgrade their on-site IT services without breaking the bank. With prices starting at less than £1,500, this compact tower server takes everything we liked about the PowerEdge T350 and boosts performance with support for Intel's Xeon E-2400 CPUs and fast DDR5 memory.

We're pleased to see Dell hasn't changed the chassis design as its build quality is outstanding. And as with most of its other SMB-focused Gen16 PowerEdge servers, Dell is targeting 'near-edge' applications and offers the T360 with an optional bezel kit that incorporates a replaceable dust filter.

Storage options haven't changed significantly but a notable new feature is support for Dell's BOSS (boot-optimized storage solution) N1 card which has dual M.2 NVMe SSD slots. Dell has also spotted a growing demand for graphics and audio computing and its response is to add an option for Nvidia's entry-level Ampere A2 GPU card.

Dell PowerEdge T360 review: Superb build quality

Lifting the server onto the lab bench left us in no doubt about build quality – for its size, the T360 is surprisingly heavy. Its chassis is constructed of sturdy pressed steel panels all round and even the chunky removable side panel tips the scales at nearly 2.5kgs.

The server comes with Dell trademark honeycomb front bezel which is also of all-metal construction. Physical security is present and correct as the bezel can be key-locked shut and the side panel has a slot at the rear for a Kensington lock.

Inside, you'll be greeted by an equally impressive design that provides easy access to all components for upgrades and maintenance. Cooling is handled by a 9cms diameter fan at the rear and the motherboard is completely covered by two very solid plastic air shrouds.

Our system was supplied with a quad-core 2.6GHz Intel Xeon E-2414 CPU topped off with a large passive heatsink. Below are four DIMM slots so the 16GB we received can be boosted to the maximum 128GB supported by the E-2400 CPUs.

Dell PowerEdge T360 review: Storage and expansion 

Storage options are reasonably good for a small business server with the front panel supporting four or eight LFF hot-plug drive bays. As with the T350, an SFF drive cage isn't available so the smaller HDDs and SSDs need to be mounted in hybrid adapters which are included with each device when you order them. 

RAID options start with the server's embedded PERC S160 controller which offers software-managed stripes, mirrors, RAID10, and 5 arrays for SATA drives. Three hardware storage controllers are available for the T360 and our system came with a PERC H355 card which offers stripes or mirrors for SAS and SATA drives. 

The non-RAID HBA355i adapter provides simple pass-through services for storage devices while the H755 adds RAID5, 6, 50, and 60 arrays along with 8GB of DDR4 cache memory. The server also comes with an embedded dual-port Gigabit LOM (LAN on motherboard) and there are four expansion slots provided for more of Dell's multi-port Gigabit or 10Gbe cards with the second slot the faster PCIe Gen5 variety.

The small bay at the front looks after Dell's BOSS-N1 card which presents two hot-plug carriers and the lockable bezel protects them from wandering fingers. We were supplied with a single 960GB M.2 NVMe SSD but you can add another or choose smaller (and cheaper) 480GB SSDs and mirror them for essential redundancy.

Along with the GPU card and 10GbE adapters, the BOSS-N1 requires a high-performance fan kit mounted behind the front panel with Dell warning that its acoustics are not recommended for typical office use. From our experiences, we'd say the extra fan noise is noticeable but certainly not intrusive – unless you're in a library.

Dell PowerEdge T360 review: Server management 

There are no compromises for server monitoring and management as the T360 is endowed with a full-fat iDRAC9 controller and dedicated Gigabit port. Its web console presents an impressive amount of information on component status and health plus a full hardware inventory and direct access to BIOS and RAID configurations.

It's worth checking out all the different iDRAC9 licences as there's no point paying for features you don't need. The free Basic license may be all a small business requires as it provides web access, hardware monitoring, and remote power controls with Express adding real-time power and temperature graphing along with email alerting. 

We prefer the Enterprise 16G license as this enables virtual media services and full OS remote control. The DataCenter 16G license is aimed at larger businesses as it adds telemetry streaming of hardware metrics for advanced analytics.

We can monitor and configure all our PowerEdge servers from one web console as we run Dell's OpenManage Enterprise (OME) server as a Hyper-V VM in the lab. After it discovered the T360's iDRAC9, we used the OME console to manage and monitor it, directly access the controller's web console, and run remote control sessions.

Import an Enterprise Advanced license to the server's iDRAC9 and you can use the OME Power Manager plug-in to view graphs of power consumption for individual components. Servers with a valid support contract can also take advantage of Dell's CloudIQ service which provides machine learning, server telemetry and predictive analytics along with a cybersecurity feature that alerts you to misconfigurations that could present a risk.

Dell PowerEdge T360 review: Is it worth it?

The PowerEdge T360 is a great choice for SMBs and remote offices looking for a powerful tower server at an affordable price. Support for the entire family of Xeon E-2400 CPUs makes it easy to customize it to your budget, it offers a good helping of fast DDR5 memory plus a decent storage potential, and busy offices will appreciate Dell's legendary build quality.

Dell PowerEdge T360 specifications

Microsoft SQL Server is one of the world's most popular database platforms. According to some estimates as many as one in three databases runs on SQL Server, alongside competitors such as Oracle, MySQL, and PostgreSQL.

However, SQL Server is not a single product. Since its original launch in 1989, there have been more than 20 major releases of SQL Server. At the time of writing four different releases are still supported, namely: 

SQL Server 2016;

SQL Server 2017;

SQL Server 2019; and 

SQL Server 2022. 

All supported releases of SQL Server can handle everyday database tasks, but there are several reasons why it's important to know which specific version you're using.

How to check your SQL Server version

Support lifecycle

If you're planning a deployment, or scoping out future database projects, you need to know that your SQL Server installation will be supported for the required duration, or you might miss out on important security or functionality updates.

Each version of SQL Server receives around five to six years of mainstream support after its release, meaning it may receive new features and performance improvements alongside security updates. After this, there's a five-year period of extended support when no new features are added, but security updates and paid technical support are still available. Once extended support ends, no more assistance or updates will be provided, even if critical security vulnerabilities are discovered.

Here are the key support dates for the current major editions of SQL Server:

As you can see, the 2016 and 2017 editions of SQL Server are already out of mainstream support and will be entirely unsupported within three years. For this reason, it's important to keep track of the SQL Server version, whether you're beginning a new project or maintaining an existing database. 

Feature availability

Each subsequent release of SQL Server brings new capabilities that weren't available in previous editions. For example, SQL Server 2016 introduced enhanced encryption capabilities, for working with sensitive information such as financial data; it also brought JSON support, making it easier to programmatically extract and process data. 

New features in SQL Server 2017 included new string manipulation functions and automatic performance tuning, while SQL Server 2019 brought support for big data clusters and improvements for parallel processing. The latest SQL Server 2022 release has many features designed to work better with Azure and other cloud data sources.

If you're building a new database application, you'll want to know which version of SQL Server you're using, to ensure you have access to all the capabilities you might need.

Security and feature updates

As well as knowing which major SQL Server edition you're using, it's important to check the minor version number (the digits after the first decimal point) to ensure that the latest available updates to security and functionality have been installed. 

An up-to-date installation might have two different version numbers, depending on how server updates have been configured. If the server is set to receive GDR (General Distribution Release) updates then only security fixes will be added to the original installation, and the version number will reflect that. If it's on the CU (Cumulative Update) branch it will also receive new features and usability enhancements, and will usually have a higher version number.

You can check the latest GDR and CU version numbers on Microsoft's SQL Server update page. At the time of writing the latest updates for the past three releases are:

SQL Server 2016 and earlier major releases were updated with milestone Service Packs as well as individual GDR and CU updates; see the Microsoft website for details of the latest version numbers for different Service Pack levels.

SQL Server version number

The moment you've been waiting for: how to check your SQL Server version number.

There are numerous methods you can use to check the major and minor version numbers of a SQL Server instance. Here are four quick ways to find the information:

1. Use SQL Server Management Studio

Microsoft's free SQL Server Management Studio tool can provide extensive information about a SQL Server installation. After you've connected the application to your SQL Server instance, check the Object Explorer window and you should see the server's version number shown in brackets after its name. If you can't see the full version number, or want more information, right-click on the server name and select Properties to open a window with many more details about the version and capabilities of the server.

2. Query the server directly

You can query a SQL Server's version number directly from a Windows command prompt. To connect to the server, enter sqlcmd -S servername\instancename, replacing server\instancename with the appropriate details for your server. If your connection is successful, you should now see a "1>" prompt: enter "select @@version", then at the "2>" prompt, enter "go". This should return a copyright statement which includes the server version number and build date.

3. Examine the executable

If you have desktop access to the computer your SQL Server instance is running on, you can check the version number by inspecting the executable file that runs the server. By default you'll find this file in a folder called C:\Program Files\Microsoft SQL Server\MSSQL16.MSSQLSERVER\MSSQL\Binn (the precise path will be slightly different for older versions of SQL Server); navigate here in the Windows File Explorer, then find the file called sqlservr.exe, right-click on it and select "Properties". In the Properties window that opens, click on the Details tab to see information about the server, including its version number.

4. Check the log files

You can also find your SQL Server version number by inspecting its logs. By default, these are stored in a folder called C:\Program Files\Microsoft SQL Server\MSSQL16.MSSQLSERVER\MSSQL\Log (again, this will be slightly different for older versions of SQL Server). Note that this folder is protected, so you'll need to either edit its security settings to give yourself access or navigate to it from an elevated Command Prompt.

Once you've got access to the Log folder, open the text file called ERRORLOG, or any older file with this name and a numbered extension. The first line of each log file contains the version number of the server.

HPE's ProLiant DL345 Gen11 is a 2U single-socket (1P) rack server with a few surprises up its sleeve. Along with support for AMD's core-heavy Gen4 EPYC CPUs plus fast DDR5 memory, its storage potential makes it a great choice for businesses with big virtualization and capacity plans that want to avoid the expense of a 2P server.

Its predecessor, the DL345 Gen10 Plus, had plenty to say in the storage department, but the DL345 Gen11 adds new mid-tray and rear cage options that boost the number of bays to a maximum of 20 LFF or 34 SFF drives. There's more as HPE also offers new front cage options that support up to 36 EDSFF E3.S NVMe SSD storage devices.

With so much storage to play with, HPE is naturally aiming the DL345 Gen11 at data protection services. Prices for base models start at a shade under £4,400 with HPE's Veeam Backup and Recovery Solution upping this to around £10K and delivering 20TB of usable capacity plus a 1-year, 10 workload subscription to Veeam's Data Platform Foundation.

HPE ProLiant DL345 Gen11 review: Storage options

HPE pulled out all the stops for our review system as it supplied it with 32 SATA SFF SSDs. The standard 8+8+8 drive bay box configuration presented 24 drives at the front and we had the midtray fitted which added another eight drives.

The space above the dual PSU bay can be used for HPE's optional stacked dual SFF cage and the tri-mode backplanes for all configurations support SAS, SATA, and NVMe devices. For LFF drives, you have room for up to twelve at the front, and four in the mid-plane and if you're prepared to sacrifice some PCIe expansion slots, you can add another four at the rear.

The motherboard has four integrated SATA ports but the embedded SR100i RAID controller found in the DL345 Gen10 Plus is no longer present, so you'll need to specify one of HPE's SR (SmartRAID) or MR (MegaRAID) storage controllers. There are plenty to choose from and our system came with HPE's top-dog SR932i-p Gen11 tri-mode PCIe card which supports 6G SATA, 24G SAS, and 16G NVMe drives plus virtually all known RAID array types and has 8GB of battery-protected cache memory.

The server supports HPE's NS204i-u Gen11 boot optimized storage device which is worth considering if you want all your standard bays free for data storage. Slipping neatly into a small dedicated bay above the PSUs, it presents two hot-plug M.2 NVMe SSDs at the rear and provides mirrored redundant storage for running an OS or hypervisor.

HPE ProLiant DL345 Gen11 review: Build quality and expansion

Removing the lid reveals the mid-tray storage shelf sitting on top of the centrally-mounted CPU heatsink. The mid-tray bays are fully hot-swappable as you lift up the levers on each side and the whole cage flips up and presents all eight for easy access.

It's simple to get at the memory slots below as you pull the sprung retaining pin out from the right side of the mid-tray which releases it for removal. HPE has redesigned the server's memory configuration as the number of DIMM slots has been reduced from sixteen to twelve. These support up to 1,152GB using 96GB DDR5 Smart Memory modules or 3TB with 256GB octal-rank 3DS modules although the latter are very expensive and you can't specify the midtray if you use them.

Cooling is handled by a bank of six hot-plug fans with CPUs up to 200W requiring the standard models. Our AMD EPYC 9124 has a 200W TDP but as our server also has the storage mid-tray, it came with the high-performance heatsink kit and fans.

Expansion potential is excellent as you have dual three-slot risers at the back that support full-height, half-length PCIe 5 cards. There's room for more as underneath the risers are two OCP 3 PCIe 5 slots which support HPE's 'o' RAID cards and a good range of network adapters with our system kitted out with HPE's quad-port Gigabit module.

Another valuable feature is improved support for graphics acceleration as you can specify up to four single-width or two double-width GPUs. HPE's front enablement kit provides four or two full-height, full-length PCIe 5 slots and leaves enough room in between for a central drive cage for twelve EDSFF E3.S SSDs or eight SFF drives.

HPE ProLiant DL345 Gen11 review: Remote management services

HPE offers an impressive range of tools, utilities, and cloud platforms for remote server monitoring and management. It all starts with the server's embedded iLO6 controller which delivers a web interface packed with status information on all critical components and provides essential security features such as Secure Start, 'silicon root of trust' firmware fingerprinting, and SPDM (Security Protocol and Data Module) hardware component and PCIe expansion card authentication.

The iLO6 links up with HPE's OneView app which we run in the lab as a Hyper-V VM for monitoring and managing HPE servers, hypervisors, storage arrays, pools, and switches. Next is HPE's iLO Amplifier Pack which offers inventory and compliance reporting services for up to 10,000 Gen8, 9, 10, and 11 servers and sends diagnostics and telemetry information to HPE's InfoSight cloud service. 

HPE's star player is its GreenLake Compute Ops Management cloud service. This provides a single portal for viewing your entire infrastructure, centrally managing server deployments, and automating system lifecycle management and firmware compliance policies.

HPE ProLiant DL345 Gen11 review: Is it worth it?

HPE's ProLiant DL345 Gen11 is an incredibly versatile 1P rack server at a very tempting price. The combination of core-heavy AMD EPYC CPU and DDR5 memory make it a cost-effective virtualization platform while its outstanding expansion potential will appeal to businesses looking for a rack-dense high-capacity storage solution.

HPE ProLiant DL345 Gen11 specifications

An FTP server is a computer that can receive, store, and transmit files over a network connection. The name stands for "file transfer protocol", a global standard for sending and receiving digital data. In this article, ITPro will break down everything you need to know about FTP servers.

What is an FTP server used for?

An FTP server can be used in any scenario where users need to store and access files remotely. In practice, FTP is mostly used for sending items over the internet, but it can also be used over a local network.

Depending on the needs of the user and the server operator, an FTP server can be configured in a few different ways. It may allow remote users to log in and upload files to their own private space, or to shared folders, similar to a cloud storage account. Alternatively – or simultaneously – it may allow "anonymous" connections so that anyone can upload and download files from specific areas.

It's important to note that the FTP server's only job is receiving, storing, and transmitting data. Connected users can upload, download, copy, rename, move, or delete remote files, but they cannot run programs or edit content over an FTP connection. However, uploaded files may be processed by other applications that have access to the server space.

Who might run an FTP server?

FTP is ideal for any organization wanting to offer simple file access for remote users. For example, a business might allow homeworkers to download resources from company servers via FTP.

Anonymous FTP is also widely used for software distribution. For example, you might use a web browser to explore an online catalog of open-source programs; when you click to download an installer, the file will be sent from an FTP server.

Behind the scenes, FTP is also widely used for website development and management. Web content is developed and tested on local machines, then uploaded via FTP to the live site; a separate HTTP server then takes care of publishing the files online.

How do I operate an FTP server?

To host an FTP server, all you need is a suitable network-connected computer running appropriate FTP server software. Such software is available for all major desktop and server operating systems; however, it's normally disabled by default and must be manually enabled and configured. This might involve setting up FTP-specific user accounts and assigning storage locations and quotas.

If you don't have access to a server of your own, you can also lease FTP space for your personal use. Most web hosting providers include FTP access with their standard packages.

Remote users wanting to access an FTP server must do so using a client application. Most desktop platforms come with command-line FTP client capabilities, but for a more intuitive and flexible experience, you can use a third-party graphical FTP client such as the cross-platform FileZilla client or WinSCP for Windows.

It's also possible to download files via FTP in a web browser, by following a hyperlink that starts "ftp://" rather than "http://:".

What are the advantages of FTP over other types of network connection?

There are several reasons why a server operator might choose to use FTP for file access, rather than other options such as WebDAV or Active Directory.

For a start, FTP is a simple, lightweight protocol that takes minimal resources to run. And because FTP users can't run programs on the server, the potential security risks are limited – although if user credentials are compromised (see below) then an FTP connection could still be used to download confidential materials.

FTP is also robust and flexible. There's no limit to the maximum size of file that can be transferred, and most graphical clients allow you to queue long lists of files to be uploaded or downloaded. If a transfer is interrupted, the client can reconnect and resume uploading or downloading as many times as necessary to send or receive the whole file. This makes FTP ideal for moving or backing up files and folders of any size and type.

What's the difference between active and passive FTP?

An FTP connection can be established in two ways, known as active or passive. From the user's point of view the difference is normally academic: most FTP clients will automatically switch between modes as needed, and the features and performance available should be identical.

However, a network administrator setting up an FTP server will ideally want to ensure that it's accessible in both active and passive modes. In the default active mode, the connecting client tells the server to send files and other information to a particular virtual network port on the user's local machine.

Unfortunately, modern firewalls and routers frequently block incoming connections from external servers, causing active mode FTP to fail. Passive mode gets around this by having the client make an outbound connection to a network port on the server, which then serves as the data channel.

Is FTP secure?

The FTP standard doesn't include encryption, so if someone's spying on your network activity they could theoretically read the contents of the files you're transferring. They could even extract the username and password that are used to log into an FTP server, and thus gain unauthorized access to the server – and to other services, if the same credentials are used.

One way around this is to ensure FTP servers are only accessed over a secure VPN connection. Another option is to use an updated protocol that includes encryption; there are two popular options with very similar titles, namely FTPS and SFTP. These both work very similarly to regular FTP, while adding encryption methods similar to secure web browsing, to ensure that usernames, passwords, and transferred data can't be read by anyone who may intercept the transmission. 

FTPS and SFTP doesn't use the same server software as FTP, but the components are readily available for desktop and server OS platforms. On the client side, almost all graphical FTP applications include support for secure FTP transfers;  users may not even be aware that they're using a secure connection.

Qsan has a fine reputation for delivering affordable storage arrays and the XCubeNXT family takes this to the next level. Fault-tolerant FC and IP SANs have always been a top priority for Qsan but these new hybrid arrays add NAS share clustering services for all-round redundancy.

On review is the top-dog XN8126D which employs the same chassis as Qsan's XCubeSAN XS5326D appliance that claimed a world's first as the only one to put 26 SAS3 SFF bays in 2U of rack height. At the rear you'll find dual controllers running in active/active mode for fully transparent SAN failover.

Expansion is superb as the controllers have dual SAS3 ports and Qsan offers four disk shelves ranging from the 12-bay XD5312 to the 26-bay XD5326. The XN8126D supports a total of 546 drives allowing capacity to be pushed to an incredible 16.7PB. 

Unlike most of the blue-chips, Qsan doesn't enforce vendor lock-in so you can choose pretty much any storage device you want. Qsan's partnership with Western Digital adds even more appeal as the XN8126D supports up to four of this vendor's supermassive UltraStar Data60 or Data120 JBODs.

Qsan XCubeNXT XN8126D review: Hardware components

Each controller comes with 16GB of DDR4 cache memory which can be field upgraded to 256GB. Our controllers are powered by eight-core Intel Xeon CPUs and a quick trip to their CLI over a serial port connection tells us these are 2.2GHz Xeon D-1735TR models.

Port choices are abundant as each controller includes integral Gigabit management and quad 10GbE SFP+ iSCSI data ports. Two module slots per controller are provided and Qsan offers dual RJ-45 10GbE iSCSI, quad 10GbE SFP+ iSCSI, dual or quad 16Gbps Fibre Channel (FC), dual 25GbE iSCSI and dual 32Gbps FC.

Hardware redundancy looks good as the controllers are supported by dual 850W hot-plug power supplies partnered by hot-plug fan modules located on each side. Qsan also offers optional cache protection modules that slot into dedicated bays above each PSU and come with M.2 flash memory and super-capacitors.

Qsan XCubeNXT XN8126D review: QSM 4 features

Qsan's XFinder app discovers the appliance on the network and provides a direct link to the QSM 4 web console. This OS has been created specifically for the XCubeNXT appliances and is very different to the QSM version used by Qsan's XCubeNAS models.

It bears more similarities with the XEVO OS used by the XCubeSAN models but its web console is devoid of any performance graphs for latency, IOPS, and throughput making it difficult to know how hard you're pushing the appliance. It's easy to use though, and your journey starts by configuring storage.

Our system was supplied with a full complement of 960GB Samsung PM1643a SAS3 SSDs and we used the wizard to choose drives, create storage pools, and decide on thick or thin provisioning, pool encryption, and auto tiering. RAID choices are extensive as along with all the usual suspects, you can select enhanced EE arrays which offer faster rebuild times by adding more spare drives to a pool.

When creating volumes, you choose the block option for FC and IP SAN presentation or file volumes for NAS shares. Snapshot protection can be applied to each volume and scheduled for as often as every five minutes.

Essential ransomware protection is present as WORM (write once, ready many) can be applied to file volumes. Unlike the QSM software for XCubeNAS, you don't have three policies to choose from and simply set a retention period in days, hours or minutes.

QSM 4 supports Rync backups with other compliant systems and in client mode, you can automatically send files to a remote system or retrieve them from it. Support for Qsan's XMirror service along with deduplication and compression aren't currently available but Qsan advised us they will be implemented in the next firmware version.

Qsan XCubeNXT XN8126D review: Clustering services

Qsan's clustering feature is designed to provide network link redundancy to NAS shares. It uses the same nominated Ethernet interface on each controller and assigns a virtual IP address to them.

For FC and IP SAN fault tolerance, you apply the standard method of connecting initiators to both controllers using MPIO. A valuable feature is the controllers' function in synchronous mode so hosts connected across both controllers have all MPIO links active and optimized.

Cluster creation is simple as you select the target interface on the controllers and assign a fixed IP address. Bandwidth for the two physical links is not aggregated as one link is active whilst the other remains on standby. 

The QSM 4 user manual doesn't mention this but you should ensure the IP addresses of the target interfaces are not on the same subnet as the cluster address to avoid potential conflicts. In practice, clustering works very well and to test this, we mapped a NAS share to a Dell PowerEdge Windows server using the cluster address assigned to a pair of 10GbE ports.

The web console showed Controller 1 was providing the primary link so we rebooted it from the QSM console while running a continuous Ping on the cluster IP address. We saw Ping timeouts while Controller 2 took over and in only 18 seconds, our NAS share was back in action.

Qsan XCubeNXT XN8126D review: Is it worth it?

Starting at under £11K for a diskless unit, the XCubeNXT XN8126D hybrid storage array is very competitively priced. The QSM 4 software makes light work of deployment but key features such as compression, deduplication, and system utilization stats have yet to be implemented.

That said, its active/active architecture delivers five 9s reliability and the cluster feature offers fast failover services for NAS shares. A great choice for mixed workloads, it provides a solid set of data protection features and an incredible storage potential.

Qsan XCubeNXT XN8126D specifications

Small businesses and remote offices seeking a compact and capable rack server will find Dell's PowerEdge R260 has a lot to offer. Designed to fit in the smallest of spaces, its short-depth 1U chassis offers a 24% footprint reduction over its predecessor, the R250.

Dell is also aiming the R260 at 'near-edge' applications such as warehouses, desktop deployments, small data cabinets, and retail offices. It can handle these harsher environments as Dell offers an optional bezel kit that incorporates a replaceable dust filter.

The small chassis doesn't compromise processing power as the R260 supports all the latest Intel Xeon E-2400 CPUs. Our system was endowed with a speedy 3.3GHz 6-core Intel Xeon E-2456 but businesses with a keen eye on costs can opt for one of two Pentium Gold G7400 dual-core models which brings the starting price down to around £1,400.

All Xeon E-2400 and Pentium G7400 CPUs support up to 128GB of memory which is the faster DDR5 variety. Storage sees improvements too, as the R260 is available with a hot-swap cage for up to six SFF HDDs or SSDs – the R250 only offers up to four LFF bays.

Dell PowerEdge R260 review: Internal design

Under the lid, you'll find a tidy interior with cable clutter kept to a minimum and easy access afforded to all key components for upgrades and maintenance. Behind the right-hand front grill are three cold-swap cooling fans that look after the CPU and four memory slots mounted further back.

We have the dual LFF drive cage which has a fourth fan behind it with Dell also offering an optional high-performance module. If you order the larger 6-bay SFF hot-swap cage, you'll receive two more standard fans which are also located behind it.

Expansion potential looks good as the R260 comes with two PCIe slots on each side of a riser with both supporting half-length, half-height cards. The server has an integral LOM (LAN on motherboard) presenting dual Gigabit ports and upgrade options are plentiful with Dell offering dual and quad-port copper Gigabit or 10GbE adapters.

You have a choice of 450W Platinum or 700W Titanium power supply units (PSUs) although most configurations only allow the latter to be selected. Redundant PSUs aren't supported but that's hardly surprising considering the size of the chassis as cooling could be an issue.

Dell PowerEdge R260 review: Storage decisions

Whichever drive cage you choose, you'll be pleased to know that the R260 has an embedded PERC S160 controller that supports software-managed stripes, mirrors, RAID10, and 5 arrays for SATA drives. The R250 has an older embedded S150 controller although its feature set is identical to the S160.

The S160 is configured from the server's BIOS setup menu where you choose AHCI for pass-through mode or RAID. With the latter enabled you can then configure your drives for the desired RAID array type.

If you want to use SAS drives you'll need an adapter card and Dell offers three PERC 11 models. The HBA355i provides pass-through services for SAS3 and SATA drives, the H355 adds stripes and mirrors while the top-dog H755 brings RAID5, 6, 50, and 60 to the table along with 8GB of DDR4 cache memory.

Another feature we're pleased to see is the bay next to the PSU for Dell's optional BOSS-N1 card. This is the latest version which presents two hot-plug M.2 NVMe SSDs at the rear and is ideal for running an OS or hypervisor on a fault-tolerant mirrored array without wasting any front bays. In fact, one of Dell's target applications for the R260 is near-edge virtualization.

Dell PowerEdge R260 review: Management and monitoring

It may be an entry-level server but the R260 won't be found wanting for remote management services as it sports the same full-strength iDRAC9 controller as its bigger PowerEdge brothers. It presents a smart web console offering a wealth of information on system and component status, power usage, and cooling efficiency along with a hardware inventory plus direct access to BIOS and storage configurations.

We ran Dell's OpenManage Enterprise (OME) software in the lab as a Hyper-V VM and after it discovered the server's iDRAC9, we could manage and monitor it, control power, and run remote control sessions. Plenty of plug-ins are available with the Power Manager activated with an OME Enterprise Advanced license applied at the iDRAC console.

This plug-in adds a set of graphs to the OME console device view showing everything from CPU, storage, memory, fan, and FPGA power consumption to component utilization and airflow in CFM (cubic feet per minute). You can use it to enforce per-rack power cap and thermal event-based policies to automatically throttle servers back as required.

Once the server has a valid support contract, you can integrate OME with Dell's free CloudIQ cloud-hosted service. OME functions as a collector for CloudIQ, sending telemetry and alert logs to it every 15 minutes, and you can use the CloudIQ portal to view all monitored servers, storage arrays, and switches, see system health status and alerts, pull up historical trending reports and create custom dashboards. The Cybersecurity feature will prove useful as it alerts you to a wide range of potential risks such as default or weak iDRAC admin passwords, SSH access, enabled USB ports, and having a secure boot disabled.

Dell PowerEdge R260 review: Is it worth it?

With prices starting as low as a shade over £1,400, the PowerEdge R260 is an affordable entry server for SMBs and also for larger businesses that want to push compute power closer to the network edge. Despite its compact dimensions, it offers a powerful hardware package, has room for further expansion, delivers plenty of storage choices, and won't be faulted for management and monitoring services.

Dell PowerEdge R260 specifications

AI presents an opportunity for partners to sell more on-premises hardware, according to HPE CEO Antonio Neri.

Speaking at the company’s annual Partner Growth Summit ahead of the main HPE Discover 2024 conference in Las Vegas, Neri said: 

“When you think about AI is not a monolithic workload, it’s actually a distributed workload, which is obviously very data intensive. That actually opens up the opportunity to bring more infrastructure on premises than ever before.

“The fact that AI is a data first business will drive customers to rethink their data strategy, where they store it, [and] how they maintain it in compliance.”

By offering customers a cloud-first approach that’s also based on-premises (such as that provided by HPE GreenLake), there’s an opportunity to really increase revenues off the back of the trend, he suggested.

“AI has given a triple shot in the arm to the server business,” he added.

It’s not just AI where HPE believes the opportunity lies, though. Leaning on one of the key themes of this year’s conference, sustainability, Neil MacDonald, EVP and GM of compute, HPC and AI at HPE, spoke of the need for companies to modernize their existing infrastructure, saying “there’s a lot of old, dusty server equipment out there that is ripe to refresh”.

MacDonald claimed that these elderly systems are “wasting our customers’ money”, as they’re not very efficient.

“There are over 400,000 [HPE] systems out there of over four years of age,” MacDonald said.

“Legacy systems are consuming about 66% of data center power, but they're only doing about 7% of useful work so there’s a huge opportunity for efficiency gains. One Gen11 can replace 11 Gen8 servers and consume 90% less power.”

He also pointed out that older infrastructure doesn’t have all the security features of HPE’s more recent products. All of these are reasons to encourage customers to upgrade, he said.

ITPro's Jane McCallion will be live on the ground in Las Vegas this week with rolling coverage of HPE Discover 2024. Keep tabs on all the latest news, updates, and announcements in our live blog.

Aimed firmly at content creators, media professionals, and small studios, Qnap's TBS-h574TX NASbook offers them a very particular set of skills. This diminutive desktop NAS appliance delivers an impressive hardware package with its pair of Thunderbolt 4 ports adding plenty of appeal to those engaged in 4K video production projects.

There are some surprises in the storage department as this all-Flash desktop NAS is the first we've seen to support EDSFF (Enterprise and Datacenter Standard Form Factor) E1.S drives. It offers five hot-plug bays across the front and Qnap has all the angles covered as the standard hot-plug carriers come with M.2 2280 adapters already fitted.

It's a powerful little box that's available with a choice of Intel's Gen13 Core i3 or Core i5 CPUs. We have the higher-end model on review which employs a 1.8GHz 12-core Core i5-1340PE CPU sporting four Performance cores with HyperThreading, eight Efficient cores, and an integrated Intel Iris Xe Graphics GPU.

Our appliance came with 16GB of non-upgradeable DDR4 memory while the Core i3 version gets 12GB. Bear in mind that Qnap's QuTS hero operating system (OS) is quite memory-hungry and if you opt for the cheaper version, you won't be able to use its integral data deduplication feature as this requires a minimum of 16GB.

Qnap TBS-h574TX review: Design and hardware

The TBS-h574TX looks good on the desktop as it's encased in a sleek and solid aluminum case with a posh gold Qnap logo on the upper surface. Weighing in at 2.24kgs unpopulated, it's easy enough to transport and the external power supply is pleasingly small as well.

The drive carriers are accessed from behind a magnetic front grill which has a small locking tab to one side. No tools are required to fit M.2 SSDs as you slot them in and let the retaining tab snap back to secure them but there's work to be done for EDSFF E1.S drives.

It depends on the type you choose as for 5.9mm high SSDs, you replace the M.2 adapter with the drive while the thicker 9.5mm and 15mm models dispense with the carrier and are screwed directly to the front locking assembly. Note that you won't see the full performance benefits of PCIe Gen4 NVMe SSDs as the appliance's slots are only the PCIe Gen3 x2 variety. Qnap recommends putting its heatsinks on them which are not included and we suggest you use them as during our performance tests, the two rear system fans ramped up and became quite intrusive.

You'll find Thunderbolt 4 ports fore and aft along with a choice of USB 3.2 and USB 2 ports. The rear panel offers an HDMI 1.4b port for piping 4K 30Hz video out from the GPU and network connections are plentiful as you have 2.5GbE and 10GbE RJ-45 with both the multi-Gigabit variety.

Qnap TBS-h574TX review: QuTS hero features

For testing, we fitted 480GB Kingston M.2 NVMe SSDs in the first two bays for use as a mirrored system pool and added two more 800GB M.2 NVMe SSDs for general data storage. Installation is a breeze as the web-based wizard disabled the default admin account, insisted we create a new one, and then proceeded to load the OS.

You can install Qnap's more nimble QTS software but QuTS hero offers a lot more features including stronger data protection features. ZFS copy-on-write delivers near unlimited NAS and iSCSI LUN snapshots along with inline data deduplication, compression, and end-to-end checksums that handle the self-healing of data corruption.

Although not so relevant in a five-bay appliance, RAID options are more extensive and you can opt for triple parity RAID to protect against three drive failures or triple mirroring which stores identical copies of data on three drives. And if you don't want your NAS data tampered with, QuTS hero offers a choice of two WORM (write once read many) policies that can be applied during share creation.

The Thunderbolt 4 ports support peer-to-peer connections up to 20Gbps with Mac and Windows hosts. Managed from the Network & Virtual Switch app, they also support directly connected external USB storage devices which automatically appear as network shares for authorized users.

Qnap TBS-h574TX review: 10GbE performance

As a NAS appliance, the TBS-h574TX has a very good turn of speed over 10GbE. With a share on the 800GB mirrored SSDs mapped over 10GbE to a Dell Windows Server 2022 host, we recorded top Iometer sequential read and write rates of 9.3Gbits/sec and 9.1Gbits/sec and swapping over to random operations returned precisely the same speeds.

These numbers highlight an important feature of the QuTS hero OS as its ARC (adaptive read cache) and ZIL (ZFS intent log) features are already very efficient making SSD caches unnecessary. Real-world speeds are just as good with a large 25GB test file copied between the NAS and server returning average read and write rates both of 9.1Gbits/sec.  

If you want the appliance to present IP SANs, you won't be disappointed as a 250GB iSCSI target mapped to the server over 10GbE delivered sequential and random read and write rates identical to the NAS share. The Core i5 CPU has a lot of horsepower on tap as during all our tests, utilisation never exceeded 9% so there's clearly plenty of bandwidth available for Thunderbolt 4 activities.

Qnap TBS-h574TX review: Is it worth it?

The TBS-h574TX NASbook is in a league of its own as no other vendor offers integrated support for EDSFF E1.S and M.2 SSDs along with dual Thunderbolt 4 ports. More system memory would have been nice but that said, it packs an impressive hardware package into a compact and very well-designed desktop chassis. With the higher-end model on review costing a shade under £1,500 excluding VAT, it's an affordable choice for content creators who want high-speed direct access to large video files.

Qnap TBS-h574TX specifications

AMD's Gen4 EPYC CPUs deliver some incredible core counts allowing single-socket (1P) servers to offer a serious alternative to more costly dual-socket (2P) servers. The high TDP of the top EPYC models does present a problem for low-profile 1U rack servers though, as going up to the more thirsty models usually requires expensive open-loop water cooling.

HPE's ProLiant DL325 Gen11 neatly solves this conundrum as it's available with HPE's closed-loop liquid cooling system. This means you can put any Gen4 EPYC CPU you want in it, including the 128-core 2.25GHz 360W 9754 and the 96-core 2.55GHz 400W 9684X.

There's a lot more to this server as along with a high core competence, it supports up to 3TB of HPE's DDR5 SmartMemory and has plenty of expansion potential. Storage features are equally good as you can choose from LFF and SFF form factors, SAS, SATA, and NVMe devices, and even up to twenty EDSFF E3.S SSDs.

HPE ProLiant DL325 Gen11 review: Build and design

As you'd expect, HPE's chassis built quality is exemplary, and cracking the lid reveals a very tidy interior with easy access to all key components. The CPU socket is centrally mounted and flanked on each side by six DIMM slots.

The liquid cooling kit must be specified for CPUs with TDPs equal to or more than 320W and is required for the 64-core 3.1GHz AMD EPYC 9554P in our review system as this has a default 360W TDP. The components take up barely any room as the CPU cold plate has two coolant pipes routed directly to a full-width radiator mounted between the seven hot-plug cooling fans and the drive backplane.

The cold plate has an integral pump powered by a short cable from a board connector next to the CPU socket. The pipes look reassuringly sturdy and the coolant is a mix of purified water, ethylene, and corrosion resistant additives 

Note that the liquid cooling kit has a maximum usage limitation of five years after which it must be replaced by HPE at the customer's expense. The server includes a standard three-year warranty but for extra peace of mind it would be well worth extending this to five years

HPE ProLiant DL325 Gen11 review: Management and monitoring

HPE's iLO6 controller keeps a close eye on the liquid cooling module and if it detects any problems it can issue SNMP traps and email alerts. The Power & Thermal page provides more details on the cooling fans and their speeds and although noise levels are irritating, this server will most likely be in a dedicated room.

Along with a very informative web interface, the iLO6 clearly has platform security as a high priority. Along with HPE's Secure Start and 'silicon root of trust' firmware fingerprinting, the SPDM (Security Protocol and Data Module) feature authenticates all hardware components plus optional PCIe expansion cards and alerts you to 'hostile' devices.

All ProLiant Gen10 and Gen11 servers can connect directly to HPE's Compute Ops Management (COM) cloud service. Part of HPE's GreenLake platform, it provides complete server lifecycle management and remote integration with HPE's OneView on-premises management platform.

We run HPE's iLO Amplifier Pack in the lab as a Hyper-V VM which provides discovery, inventory, and compliance reporting for up to 10,000 Gen8, 9, 10, and 11 servers. It can send inventory and performance data to HPE's InfoSight cloud service which can now be accessed directly from the GreenLake portal.

HPE ProLiant DL325 Gen11 review: Storage and expansion

For a low-profile 1U rack server, the DL325 Gen11 offers a surprisingly good range of storage options. Our system came kitted out with the eight SFF drive backplane and a full house of SATA SSDs which can be bumped up to ten with an extra front dual-bay cage.

Other storage configurations are a four LFF drive backplane or a twenty EDSFF E3.S SSD version. The server supports up to four single-width or two double-width GPUs and in these configurations, you're still left with a single bay to the left which supports four SFF drives or eight EDSFF E3.S SSDs.

AMD's chipset doesn't offer any integrated RAID features so unless you intend to use software-managed arrays, you'll need to source one of HPE's controller cards. There are plenty to choose from including the Gen11 Tri-Mode PCIe adapters and OCP 3 versions with the top-dog SR932i-p supporting 24G SAS, SATA, and NVMe devices.

Our system was also supplied with HPE's NS204i-u Gen11 boot-optimised storage device which is worth considering if you want all your front bays free for data storage. Slipping into a dedicated bay and cabled directly to the motherboard, it presents two hot-plug M.2 NVMe SSDs at the rear and provides mirrored redundant storage for running an OS or hypervisor.

There's plenty of room for further expansion as the server has two risers each sporting PCIe Gen5 x16 slots with enough space in front of the primary slot for a full-height, full-length card. Underneath these you'll find two OCP 3 PCIe 5 slots which support HPE's 'o' RAID cards and a good range of network adapters with our system supplied with HPE's quad-port Gigabit module.

HPE ProLiant DL325 Gen11 review: Is it worth it?

The ProLiant DL325 Gen11 is a good value single-socket 1U rack server and with AMD's Gen4 EPYC CPUs in the driving seat, looks set to take on a wide range of enterprise workloads. The air-cooled version is a very capable system but HPE's closed-loop liquid cooling system adds extra versatility as it allows the server to support up to 400W TDP CPUs.

The liquid-cooled model on review is a great choice for virtualization tasks that require high core densities. It will also appeal to businesses that want to squeeze maximum processing power from their rack cabinets without heavy investment in open-loop water cooling systems.

HPE ProLiant DL325 Gen11 specifications

In all but the most cloud-based businesses, setting up a home, small, or branch office is likely to mean investing in a network-attached storage (NAS) device. Choose well and you'll accelerate workflows, add a layer of data security, and solve loads of provisioning headaches – but get it wrong and you could be looking at upgrading much sooner than expected.

Asustor's Nimbustor 4 Gen2, also known as the AS5404T, is a four-bay NAS enclosure. It's aimed at the enthusiast home and small business market and is a well-specified device, equipped with a potent – in this class – Intel Celeron N5105 processor and 4GB of RAM as standard. At the rear you'll find a couple of 2.5GbE ports, along with two of its three USB 3.2 Gen2 (10Gb/sec) ports – there's a third on the front panel. The back panel also has an HDMI port, allowing you to turn this NAS into a basic workstation, surveillance station, or multimedia hub.

Unusually – possibly uniquely – this NAS supplements its four SATA drive bays with four NVMe slots for SSDs. While other appliances in this class tend to offer two slots, having four opens up more possibilities – such as using two SSDs for a read/write cache and two more to provide a fast storage pool with data redundancy.

The AS5404T certainly allows for more expansion than you'd get from a typical two-bay device. For example, you could start with two disks and two NVMes in RAID1 arrays, then add disks later to convert one or both to RAID 5, RAID 6, or RAID 10. If you need more storage, you could also add up to three Asustor AS6004U expanders via USB, giving a maximum total of 16 disks and four NVMes. While there's no PCI expansion slot for upgrades such as 10GbE networking, you can aggregate the two network connections to double the overall network throughput.

Asustor Nimbustor 4 Gen2 review: Setting up

The AS5404T is sold without disks or NVMes. It's easy to add them, and the process is tool-less for 3.5" SATA drives. Screws are provided to support 2.5" disks, so you'll obviously need a driver to fit those. While this NAS' magnetic front cover comes off easily, its caddies release with a really unpleasant cracking sound that sounds like you've just broken a latch. We never quite got used to it.

It's a bit more involved to fit NVMes, as you'll need to remove one side of the plastic casing. With this done, the slots are easily accessed, and the SSDs simply clip into place. Two memory slots are accessible from within the disk bays, although you'll need to remove the fitted 4GB SODIMM to upgrade to the 16GB maximum. All told, this isn't the hardest NAS to set up – you could fit four disks, four NVMes, and some extra memory in less than 10 minutes.

With the physical setup finished, you can find the NAS on the network with the Asustor Control Center app, or just check your DHCP server to find which IP address to browse. This NAS supports the usual four-bay RAID options for both disks and SSDs, and either EXT4 or Btrfs formatting. Once you've installed and configured Asustor's ADM operating system, it's quite easy to configure disk and SSD volumes through Storage Manager, or to set up SSD caching of the main disk volume. One potential limitation is that you can't accelerate multiple disk volumes with a single SSD cache.

It's quite easy to connect up both of this NAS' network ports and create an aggregated link. This feature supports seven different configurations, which variously offer fault tolerance or load balancing. There's also 802.3ad aggregation, although you'll need a compatible switch for that. While load-balancing could roughly double this NAS' maximum network bandwidth, individual users won't get transfer speeds above the 2.5GbE maximum of around 283MB/sec.

Asustor Nimbustor 4 Gen2 review: Features and apps

Like most other NAS operating systems, ADM is Linux-based. While it's not quite as slick as Synology's DSM, it's still a relatively straightforward OS, in which you'll easily find the bulk of features you might want for a small office. ADM is arguably a bit light when it comes to disk and volume management, particularly when compared to QNAP devices that offer a choice of conventional, thin, and thick volumes. Still, it does enough for the vast majority of users, and it's generally a little easier to use.

ADM excels when it comes to apps, however. At the time of writing, 272 were available in App Central – more than twice as many than either the Synology Package Center or QNAP's App Center offer on similarly priced devices. Business users are well served, with apps spanning communication and content management, including Joomla and WordPress, along with various database, backup, media serving, and surveillance software. As you'd expect, you can configure remote access and management through EZ-Connect. There's also VPN support through Asustor's own VPN Server app. As with other NAS boxes, most of the software is free. 

Asustor Nimbustor 4 Gen2 review: Performance

Given this NAS' flexibility, we wanted to test it in a few configurations. Initially, we loaded up four 1TB disks, and two 1TB SSDs loaned to us by Western Digital. We configured the former as a RAID 5 array and the latter as a stand-alone RAID 1 volume, and waited until the arrays had finished synchronizing. Connected via a single 2.5GbE port, we then used ATTO Disk Benchmark to record the read and write performance across a spread of file sizes.

The AS5404T was fast, managing to hit nearly 48MB/sec read and 45MB/sec write speeds for 4KB operations, and reaching 164MB/sec reads and 167MB/sec writes for 16KB operations – among the fastest results we've measured on this class of NAS. For 2MB operations, it could read at 281MB/sec – almost the 2.5GbE ceiling – although its 256MB/sec write speeds were slightly more prosaic.

To check whether we'd get the same performance in the real world, we timed file copies to and from a mid-range Windows 11 laptop. Here the NAS comfortably maintained 283MB/sec for a single, multi-gigabyte file. As we would expect, throughput was much slower for smaller files, reducing the overall performance across our 11,000 file dataset to a 71MB/sec read/write average – again, that's one of the fastest we've measured in this class.

Perhaps surprisingly, there was little change in this NAS' performance when we tested its SSD volume, nor when we reconfigured it with two hard disks in a 12TB RAID 1 array. Less surprisingly, we saw no benefit from SSD caching, which typically becomes more effective over time in a multi-user environment.

We used the Plex app to set up a media server and separately started a large transcoding job on Handbrake, which here runs in a Docker container. With the CPU at nearly 100% and a large file copy running, HD movie playback was smooth and responsive, suggesting the Nimbustor 4 Gen2 would cope well with a reasonably demanding workload.

Asustor Nimbustor 4 Gen2 review: Is it worth it?

For the money, this is simply an excellent NAS. While it doesn't have the greatest operating system, and we came to hate its snappy drive caddies, it offers the perfect balance of power and features. It might not have the most businesslike appearance, but it's a great choice to provide data storage, and useful services and utilities in a small office. Best of all, its flexible drive support and upgradeability means it could grow with your business, so you won't be rushing to upgrade if things really take off.

Asustor Nimbustor 4 Gen2 specifications 

Type

Barebones NAS enclosure

Drive support

4 x SATA, 4 x NVMe

RAID support

RAID0, RAID1, RAID5, RAID6, RAID10, JBOD

Dimensions (HWD)

170x174x230mm

Processor

Intel Celeron N5105

Memory (MAX)

4GB (16GB)

Network

2 x 2.5GbE

Other ports

3 x USB 3.2 Gen2, 1 x HDMI

Warranty

Three year RTB

The Asustor Flashstor 6 brings all-flash storage to the affordable NAS market. It's really aimed at home power users, who may be happy to sacrifice outright storage space in exchange for fast file access and powerful media serving capabilities. However, we wanted to know if it might also provide a cheap all-flash alternative for demanding small businesses. We're testing the FS6706T, which has six NVMe slots, but the Flashstor 12 (FS6712X) is essentially identical while offering twice the storage. Both support disk configurations including RAID 5, RAID 6, and RAID 10.

This NAS has a reasonable all-round specification. It's based on an Intel Celeron N5105 processor which, while not the fastest you'll find in a NAS, should have the grunt to cope with fairly demanding use in a small office. It's paired with 4GB of memory which you can upgrade to a 16GB maximum, although this means opening up the top cover and parting company with the SODIMM already installed. At the rear, there are dual 2.5GbE network ports, two USB 2 ports, and a USB 3.2 Gen2 (10Gb/sec) connector. There's also one of these on the front panel.

This NAS' rear panel is completed with a 4K-capable HDMI port, opening up the possibility of using it as a workstation, surveillance center or media center – there's an S/PDIF audio port to help with this, too.

You can't fit conventional disks to this NAS, but you can use its USB 3.2 ports to connect up to two of Asustor's AS6004U four-disk storage expanders. There's no flash storage expansion unit yet, so if you might want to add more SSDs it's wise to consider the FS6712X instead. Disappointingly, this costs about £250 more, which seems quite a premium for another six slots.

Asustor Flashstor 6 review: Set up

It's worth mentioning the Flashstor 6's unusual form factor. Despite the potential space savings from using NVMes, Asustor has chosen to flatten this NAS, leaving it with a footprint roughly the same as a compact six-bay SATA enclosure. It looks much like a home Wi-Fi router, even down to its prominent LED lighting – you can use the night mode feature to tone this down, but it's likely to upstage drab printers and any other NAS enclosures nearby.

While populating a NAS with SATA disks is usually tool-less, you'll still need a screwdriver to get NVMes into most examples. The FS6706T is no different – its slots are found behind a cover on the underside that's secured by four screws. Each disk clips into place, after which you can refit the cover and get started.

It's fairly quick and simple to install Asustor's ADM operating system. You can use the Asustor Control Center app to locate the NAS from a PC, or find the IP address it's leased and connect to the web admin interface that way. We had the NAS up and running within about 15 minutes, including the time taken to install two NVMes. While ADM still defaults to the EXT4 file system, you can choose Btrfs instead.

Asustor Flashstor 6 review: Features and apps

Asustor is marketing the FS6706T at creatives and gamers, and there's no doubt it's well suited for the role. It looks like a home device and is designed to excel at taxing jobs like real-time 4K H.265 transcoding and media serving. The company makes the point that, if using iSCSI, the FS6706T's storage appears to a PC as a local drive, which should also allow users to install games to the NAS, rather than the limited flash storage on a gaming rig.

But while not specifically aimed at business, the Flashstor 6 has most of the tools and features the typical small business is likely to need. You can set it up as a DHCP server, configure remote access, and use various options to synchronize or backup data to and from other servers, for example.

In addition, there's a huge range of software in App Central – we counted 274 packages at the time of our review. Many of these are centered around streaming, content production, and content management, but there's also a wide range of utilities including VirtualBox, Asustor Surveillance Center, and a choice of SQL servers. Importantly, you'll also find Docker Engine, which lets you run containerized apps and add even more functionality.

Asustor Flashstor 6 review: Performance

This NAS has compatibility with a wide range of NVMe SSDs, but we only had two suitable ones available at the time of our review. That limited us to RAID 0 or RAID 1 configuration – we chose RAID 1 for data redundancy. We connected to the NAS via a single 2.5GbE connection and used ATTO Disk Benchmark to measure read and write performance across a range of file sizes.

The results were quite impressive. Even with our two-disk array, the FS6706T managed 45MB/sec writes and 48MB/sec reads for 4KB operations. At 16KB these speeds increased to 169MB/sec and 165MB/sec respectively, while at 2MB this NAS was writing at 249MB/sec. At nearly 283MB/sec, its 2MB read performance was virtually at the maximum possible over a single 2GbE connection.

We always check a NAS' data speeds with a real-world Windows copy, which in this case involved an 11GB data set comprising 11,000 files. These broke down into a single 8GB file, 2GB of photo files typically a couple of megabytes each, and 10,000 very small files. The FS6706T had no trouble reading and writing the single file at 283MB/sec, and it managed a very impressive 95MB/sec when writing the entire set. Curiously, it was slower when reading smaller files, reaching just 41.5MB/sec over the whole set, to give a 68MB/sec read/write average – still one of the quicker performances we've measured in this class of device.

It's worth pointing out that our test configuration wouldn't have provided this NAS the opportunity to deliver its best performance. We'd expect a six-SSD RAID 5 array to perform more quickly, potentially giving this NAS a significant edge over disk-based alternatives. This should be especially the case for smaller files and multiple concurrent users. It's also worth pointing out that, by aggregating both network ports, you could expect greater throughput in a multi-user environment, even though individual connections would be limited to the 283MB/sec maximum possible via 2.5GbE.

We finished our tests by configuring a transcoding job in Handbrake, starting a large file transfer in Windows, then streaming an HD movie via Plex. Despite the processor running almost flat out, video playback was completely smooth. Although the fan speed stepped up slightly, this remained a very quiet NAS, particularly as there's no additional hard disk noises or vibration to bother you.

Asustor Flashstor 6 review: Is it worth it?

Clearly, the Asustor Flashstor 6 isn't built for demanding data center applications. It misses out on enterprise features like data deduplication, that serve to prolong the life of heavily worked SSDs, while its 2.5GbE interfaces simply won't do for the most intensive data access. But that's to miss the point. As a more affordable all-flash device for data-heavy small offices, it provides a surprising amount of performance for the money.

That said, it's hard to make the case for the FS6706T as a single, do-it-all NAS. It's well-served for features and, with 8TB SSDs in RAID 5, would offer up to 40TB of fast network storage. However, Asustor's own AS5404T combines four NVMe slots with four SATA bays and can scale up to 16 disks with compatible expansion units. The Flashstor 6 is a good choice if you want a faster NAS to supplement existing disk-based storage, but if you're starting from scratch, the AS5404T is a more scalable, flexible device – and it performed almost identically in our tests.

Asustor Flashstor 6 specifications