RESEARCH TRIANGLE PARK – If “Terminator” were written today, why not put the artificial mind in the “cloud” where there’s no single point of failure, let along a “bare metal” host site subject to discovery and destruction? Such a Skynet creation could possibly survive damage unless the entire cloud infrastructure – and with it most if not all of the internet – were destroyed.

Not going to happen unless some “doomsdayers” decide to take man back to analog. Perish the thought!

Which brings us to Big Blue – not Big Brother – and its move to take artificial intelligence into the cloud minus all the hardware.

Yes, IBM (and let’s not leave out Red Hat, IBM’s core cloud player) has found another way to tout its cloud computing business by creating what it calls an artificial intelligence-focused supercomputer that exists in the cloud.

Rick Smith is editor and cofounder of WRAL TechWire

Called Vela (plural of the Roman word velum for sail), its existence – maybe a Skynet in the making since its AI? – the effort was unveiled by IBM Research in a blog post.

It’s intriguing, to say the least, given all the hype and frenzy around the emergence of AI-enabled chatbots such as ChatGPT. And the news comes in transformative times for IBM under Chair and CEO Arvind Krishna.

Although operational since May of 2022, Vela hasn’t been discussed much by Big Blue. Discussing it now makes sense as investment dollars flood into tech AI startups and Microsoft as well as Google prowl for AI opportunities. An interesting play: go cloud without lots of “bare metal” hardware – but why?

“Having the right tools and infrastructure is a critical ingredient for R&D productivity. Many teams choose to follow the ‘tried and true’ path of building traditional supercomputers for AI,” the IBM researchers wrote. “While there is clearly nothing wrong with this approach, we’ve been working on a better solution that provides the dual benefits of high-performance computing and high end-user productivity, enabled by a hybrid cloud development experience.  Vela has been online since May 2022 and is in productive use by dozens of AI researchers at IBM Research, who are training models with tens of billions of parameters.”

AI wars: These players are fighting for dominance among artificial mind chatbots

The researchers point out that before Vela “building a supercomputer has meant bare metal nodes, high-performance networking hardware (like InfiniBand, Omnipath, and Slingshot), parallel file systems, and other items usually associated with high-performance computing (HPC). But traditional supercomputers weren’t designed for AI; they were designed to perform well on modeling or simulation tasks, like those defined by the US national laboratories, or other customers looking to fulfill a certain need.”

Inside Vela

The IBMers point out what they see as cloud advantages.

In discussing Vela, IBM says “traditional design point has historically driven technology choices that increase cost and limit deployment flexibility. We’ve recently been asking ourselves: what system would we design if we were exclusively focused on large-scale AI?

“This led us to build IBM’s first AI-optimized, cloud-native supercomputer, Vela. It has been online since May of 2022, housed within IBM Cloud, and is currently just for use by the IBM Research community. The choices we’ve made with this design give us the flexibility to scale up at will and readily deploy similar infrastructure into any IBM Cloud data center across the globe.  Vela is now our go-to environment for IBM Researchers creating our most advanced AI capabilities, including our work on foundation models and is where we collaborate with partners to train models of many kinds.”

IBM transformation continues: 30th buy in CEO’s brief tenure is software startup

Other AI supercomputer developments

IBM is not alone in building AI supercomputers, however.

“Last year, Microsoft unveiled its own platform using Azure infrastructure combined with Nvidia’s GPU accelerators, network kit, and its AI Enterprise software suite. This was expected to be available for Azure customers to access, but no time frame was specified,” tech news site The Register points out. It says Meta and Tesla also have AI projects but both follow  “the traditional on-premises infrastructure route, include Meta and Tesla.”

For more details, read the full blog post.

As for Skynet, maybe James Cameron will revisit doomsday after he wraps up the Avatar franchise.


Rick Smith is cofounder and editor of WRAL TechWire.