The CPU is designed for AI infrastructure and high-performance computing.
NVIDIA has unveiled their new discrete data center CPU designed for AI infrastructure and high-performance computing, providing the highest performance and twice the memory bandwidth compared to today’s leading server chips. The announcement came at GTC, NVIDIA's annual conference for AI developers.
The CPU, named Grace Superchip, has 144 Arm cores in a single socket, offering an estimated performance of 740 on the SPECrate, and complements NVIDIA’s first CPU-GPU integrated module, the Grace Hopper Superchip, announced last year. According to the announcement, Grace Superchip also provides industry-leading energy efficiency and memory bandwidth thanks to the LPDDR5x memory subsystem that offers double the bandwidth of traditional DDR5 designs at 1 terabyte per second while consuming dramatically less power, and Error Correction Code.
The company also adds that the CPU will run all of NVIDIA’s computing software stacks, including NVIDIA RTX, NVIDIA HPC, NVIDIA AI, and Omniverse.
"A new type of data center has emerged – AI factories that process and refine mountains of data to produce intelligence," comments Jensen Huang, Founder and CEO of NVIDIA. "The Grace CPU Superchip offers the highest performance, memory bandwidth, and NVIDIA software platforms in one chip and will shine as the CPU of the world’s AI infrastructure."