October rundown: AWS chaos and supercomputers surging

As the dust settled on the AWS outage, the US Department of Energy announced a slew of new supercomputers for national security

The text "AWS chaos and supercomputers surging" against a Photoshop of a jack-o'-lantern with a jagged mouth representing cloud outages. The words "AWS chaos" are in yellow, the rest are in white. In the bottom-right corner, the ITPro Podcast logo is shown.
(Image credit: Future)

It’s Halloween and what better way to mark the day than discuss one of the biggest horror stories of October?

We are of course referring to the AWS outage, which on 20 October took down some incredibly notable websites and caused global disruption. It’s the latest, most severe example of why data centers are considered critical infrastructure – and acts as a stark reminder to IT administrators that even big tech can fall foul of technical errors.

Of course, October hasn’t just been defined by negative stories. At the end of the month, the US Department of Energy has revealed a slew of supercomputer announcements. What are they for, and when can they be expected to be up and running?

In this episode Jane and Rory speak once again to Ross Kelly, ITPro’s news and analysis editor, to the show to discuss the biggest developments of the month.

Highlights

"A lot of services were down. Some of the biggest websites, applications, platforms on the web, spanning the consumer and enterprise spaces, were were impacted by this. I mean, you guys know Slack was down for a period, so we were having to communicate over WhatsApp, but major banks, Lloyds, Halifax, their apps were down and Snapchat was down."

"AWS, the engineers that were scrambling to fix this, should be commended because by late morning UK time, they had mitigated the error and resolved this. But it was the actual recovery period for a lot of organizations that was going to be quite lengthy."

"...I've written about the launch of not one, not two, but eight supercomputers, or the announcement they're going to be created. So these are quite an eclectic mix of systems.

"Now, what I will say is that of all of these, of all of them, we have a solid date for one, which is Equinox. We only have that, and it's the first half of 2026. In a press briefing from HPE, before we knew what the big announcement was going to be on Monday, we were told that the first systems on the Cray GX, 5000 would be delivered in 2027 so you can guess that Discovery and an AI cluster, so not another supercomputer, an AI cluster called Lux – both of which will be at Oak Ridge – will probably be online In 2027."

"China's top supercomputers don't rank in the Top 500 whatsoever, because it's not public information. The US and this administration, which is vocally hyping up the competition with China, talking about how the US has to maintain its foothold as the home of AI and the home of high performance computing takes the opposite approach, where they want to dominate the Top 500 list. They want DOE to have as many of those top 10 supercomputer spots as possible. But it does seem that that takes precedent over then going on to explain what they're actually being used for."

Footnotes

Subscribe 

Rory Bathgate
Features and Multimedia Editor

Rory Bathgate is Features and Multimedia Editor at ITPro, overseeing all in-depth content and case studies. He can also be found co-hosting the ITPro Podcast with Jane McCallion, swapping a keyboard for a microphone to discuss the latest learnings with thought leaders from across the tech sector.

In his free time, Rory enjoys photography, video editing, and good science fiction. After graduating from the University of Kent with a BA in English and American Literature, Rory undertook an MA in Eighteenth-Century Studies at King’s College London. He joined ITPro in 2022 as a graduate, following four years in student journalism. You can contact Rory at rory.bathgate@futurenet.com or on LinkedIn.