Meta at the moment introduced its next-generation AI platform, Grand Teton, together with NVIDIA’s collaboration on design.

Compared to the corporate’s earlier era Zion EX platform, the Grand Teton system packs in additional reminiscence, community bandwidth and compute capability, stated Alexis Bjorlin, vice chairman of Meta Infrastructure Hardware, on the 2022 OCP Global Summit, an Open Compute Project convention.

AI fashions are used extensively throughout Facebook for companies equivalent to information feed, content material suggestions and hate-speech identification, amongst many different purposes.

“We’re excited to showcase this newest family member here at the summit,” Bjorlin stated in ready remarks for the convention, including her because of NVIDIA for its deep collaboration on Grand Teton’s design and continued assist of OCP.

Designed for Data Center Scale

Named after the 13,000-foot mountain that crowns considered one of Wyoming’s two nationwide parks, Grand Teton makes use of NVIDIA H100 Tensor Core GPUs to coach and run AI fashions which might be quickly rising of their dimension and capabilities, requiring higher compute.

The NVIDIA Hopper structure, on which the H100 is predicated, features a Transformer Engine to speed up work on these neural networks, which are sometimes known as basis fashions as a result of they will tackle an increasing set of purposes from pure language processing to healthcare, robotics and extra.

The NVIDIA H100 is designed for efficiency in addition to vitality effectivity. H100-accelerated servers, when linked with NVIDIA networking throughout hundreds of servers in hyperscale knowledge facilities, might be 300x extra vitality environment friendly than CPU-only servers.

“NVIDIA Hopper GPUs are built for solving the world’s tough challenges, delivering accelerated computing with greater energy efficiency and improved performance, while adding scale and lowering costs,” stated Ian Buck, vice chairman of hyperscale and excessive efficiency computing at NVIDIA. “With Meta sharing the H100-powered Grand Teton platform, system builders around the world will soon have access to an open design for hyperscale data center compute infrastructure to supercharge AI across industries.”

Mountain of a Machine

Grand Teton sports activities 2x the community bandwidth and 4x the bandwidth between host processors and GPU accelerators in comparison with Meta’s prior Zion system, Meta stated.

The added community bandwidth allows Meta to create bigger clusters of programs for coaching AI fashions, Bjorlin stated. It additionally packs extra reminiscence than Zion to retailer and run bigger AI fashions.

Simplified Deployment, Increased Reliability

Packing all these capabilities into one built-in server “dramatically simplifies deployment of systems, allowing us to install and provision our fleet much more rapidly, and increase reliability,” stated Bjorlin.