OVHcloud just open sourced its data center liquid cooling system — here’s why

Building and logo of OVHcloud's HQ in Roubaix
(Image credit: Getty Images)

OVHcloud has announced it will be making some of its proprietary water cooling schematics open source in a bid to bolster data center industry sustainability

The French cloud computing giant made the schematics for two of its water-cooling components public during the 2024 Open Compute Project (OCP) EMEA Regional Summit, held in Lisbon on 24 April 2024.

In two dedicated whitepapers, OVHcloud’s head of R&D-Cooling, Ali Chehade detailed the technical specifications and designs of the Manual Control Valve (MVC) and 3 Balls Flow Meter.

The MVC system simplifies the hydraulic circuit, Chehade explained, using a standard commercial valve that doesn’t require any sacrifice in terms of functionality.

Chehade said the MVC’s analog design is cheaper to maintain, more reliable, and will improve operational efficiency by virtue of the solution’s simplicity.

“[A]utomatic regulation valves come with certain drawbacks, including the need for sophisticated filtering and shutoff valves for maintenance, and they are often bulky and expensive,” he said. “In contrast, ball valves are cost-effective, maintenance-free, reliable, and simple.”

In the whitepaper, Chehade emphasized that because the MVC is a manual solution the flow can be shut off at any time, this means there’s no need for a shutoff valve to be included during the maintenance process, further boosting the simplicity and efficiency of the system.

Meanwhile, the 3 Balls Flow Meter uses three balls of varying sizes along a tube in order to generate resistance to the water flow, billed as a cost-effective flow-metering solution.

OVHcloud’s 3 Balls Flow Meter solution uses push-in fittings technology to get around a number of drawbacks associated with the existing analog flow meters and flow indicator solutions.

The simplicity of the 3 Balls Flow Meter, like the MVC, is touted as its greatest strength in the OCP whitepaper, comprising just two push-in fittings, a pipe, four obstacles, and three balls.

Chehade suggested the advantages of OVHcloud’s 3 Balls flow metering solution over traditional were that it does not require sophisticated filtering and maintenance, is easy to integrate within a rack, provides precise flow readings, and doesn’t break the bank.

Like the MVC, the 3 Balls Meter is broadly applicable across a number of liquid-cooled racks across data centers, and similarly scalable. 

OVHCLoud eyes sustainability boosts in the era of energy-intensive AI

Reducing the capital expenditures for liquid-cooled data centers is a top concern as operators trying to increase capacity to match the demand for energy-intensive AI tools, while staying on track with ESG goals.

Offering cost-effective, maintenance-free solutions could be an important part of improving the sustainability of data centers, OVHcloud claimed, as global power shortages increase the pressure on cloud infrastructure providers.

To address the increased workloads placed on data centers today, liquid cooling has become increasingly prevalent in facilities as operators try to meet sustainability goals.

OVHcloud framed its decision to open up its MVC and 3 Balls Flow Meter components to the entire market as a boon to global data center sustainability efforts, as they can help simplify and optimize the water cooling systems.

By using readily available, commonly used components, the firm said its MVC can be easily integrated into existing liquid-cooled systems, allowing for scalability and applicable across a wide range of different solutions.

In a similar vein, it claimed its 3 Balls Flow Meter will lower the overall environmental impact of data center operations by reducing both water and energy consumption.

The cloud provider added that it’s also working on open source firmware via the OpenBMC project to provide update support and security guarantees for servers, commissioned or decommissioned.

Solomon Klappholz
Staff Writer

Solomon Klappholz is a Staff Writer at ITPro. He has experience writing about the technologies that facilitate industrial manufacturing which led to him developing a particular interest in IT regulation, industrial infrastructure applications, and machine learning.