Home Web graphics Nvidia is chosen to power the world’s fastest supercomputer – Facebook’s AI research data center

Nvidia is chosen to power the world’s fastest supercomputer – Facebook’s AI research data center


The company formerly known as Facebook, Metaplatforms (NASDAQ:FB), has just unveiled the details of its new AI Research SuperCluster (RSC). It’s a massive supercomputer housed in a data center, and once it’s fully built in mid-2022, it’ll be the most powerful in the world (at least as far as we know, based on announced supercomputer projects publicly).

Not surprisingly, this is a joint press release with Nvidia (NASDAQ: NVDA). The GPU designer has established itself as the leader in artificial intelligence (AI) hardware technology and has partnered with Meta in the past. This time is no different as it’s Nvidia’s latest hardware that will power Meta’s RSC.

Image source: Getty Images.

Building the Foundations of the Metaverse

The new Meta installation will be truly impressive. Equipped with 760 NVIDIA DGX A100 systems (a computing unit specially designed by Nvidia for high-end computing in data centers), it integrates 6,080 Nvidia GPUs in the supercluster, all linked together using network hardware acquired by Nvidia when it bought out Mellanox a couple years ago.

Currently, RSC offers a maximum of nearly 1.9 exaflops of performance (one exaflop equals 1,000 petaflops, supercomputing jargon for how fast the computer can perform calculations). However, a second phase of construction later this year will expand the RSC to 16,000 GPUs and increase computing power by up to 5 exaflops. I’m not sure what that means? Here is a point of reference. The Fugaku supercomputer in Japan, considered the fastest in the world today, has a maximum performance level of 2.15 exaflops.

When Facebook rebranded itself as Meta Platforms late last year, it said it was serious about building the metaverse and backed it with a pledge to spend $10 billion from it. year (and from there). The RSC is clearly delivering on that original promise. Meta says he will use his Nvidia-powered machine to research and create services using computer vision (giving a computer the ability to “see” and recognize its surroundings) and natural language processing and recognition (like the ability to have a conversation with a device, or converse with groups of people who speak different languages ​​in real time).

A boon for chip companies

The beauty of Nvidia’s hardware custom-built to handle huge amounts of data isn’t just the gear itself. Nvidia also bundles an extensive software portfolio with its DGX A100 system aimed at simplifying complex AI research and development. Nvidia’s portfolio includes image recognition software, recommender systems, and conversational AI, all ready to deploy so a developer can start experimenting and building a new software service.

Historically, supercomputers like the RSC were the domain of academia, but the work of companies like Nvidia has made AI more accessible than ever. A sort of race for AI supremacy could begin. Expect to see more announcements of mega-supercomputers like those from Meta in the coming years as other major companies attempt to develop the Metaverse (i.e. future platforms and services web-based computing) and capturing share of the ever-expanding computing industry.

What does this mean for Nvidia? Although the semiconductor industry is cyclical (periods of booming sales are often reset by periods of slow or brief sales declines), Nvidia’s overall trajectory should continue to rise for many years to come. This bodes well for other chip designers also playing in this growing AI sandbox, like Advanced micro-systems – which, by the way, currently supplies Nvidia with a pair of processors for every DGX A100 data center system built.

Long story short, it looks like Meta is betting big on shaping the future of computing with RSC, and Nvidia is playing a key role in that process. Data center sales have been a major contributor to Nvidia’s rapid growth over the past few years, and that trend looks far from over. This underscores Nvidia’s momentum behind it as the AI ​​computing arms race heats up.

This article represents the opinion of the author, who may disagree with the “official” recommendation position of a high-end consulting service Motley Fool. We are heterogeneous! Challenging an investing thesis — even one of our own — helps us all think critically about investing and make decisions that help us become smarter, happier, and wealthier.