Meta has deployed Nvidia's standalone Grace CPUs at scale and is preparing to field Vera CPUs beginning next year. Grace previously shipped mainly as part of Grace-Hopper Superchips paired with integrated GPUs, and Meta plans to deploy millions of GB300 and Vera Rubin Superchips for recommender and other datacenter workloads. Meta is now using Grace-only systems for general-purpose and agentic AI tasks that do not require GPUs, reporting roughly 2x performance-per-watt on backend workloads. Grace CPUs feature 72 Arm Neoverse V2 cores up to 3.35 GHz and support large LPDDR5x memory configurations.
"Move over Intel and AMD - Meta is among the first hyperscalers to deploy Nvidia's standalone CPUs, the two companies revealed on Tuesday. Meta has already deployed Nvidia's Grace processors in CPU-only systems at scale and is working with the GPU slinger to field its upcoming Vera CPUs beginning next year. Apart from a few scientific institutions, Nvidia's nearly three-year-old Grace CPUs have predominantly shipped as part of so-called "Superchips" which featured integrated Hopper and Blackwell GPUs onboard."
"As a refresher, Nvidia's Grace CPUs are equipped with 72 Arm Neoverse V2 cores clocked at up to 3.35 GHz. The chip is available in a standalone config with up to 480 GB of memory or two of the processors can be combined with up to 960 GB of LPDDR5x memory to form the Grace-CPU Superchip. LPDDR5x isn't something you typically see in server platforms, but offers several advantages in terms of space and bandwidth."
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]