NVIDIA Grace Hopper Superchip

The breakthrough accelerated CPU for giant-scale AI and HPC applications.

Higher Performance and Faster Memory—Massive Bandwidth for Compute Efficiency

The NVIDIA GH200 Grace™ Hopper™ Superchip is a breakthrough accelerated CPU designed from the ground up for giant-scale AI and high-performance computing (HPC) applications. The superchip delivers up to 10X higher performance for applications running terabytes of data, enabling scientists and researchers to reach unprecedented solutions for the world’s most complex problems.

Take a Closer Look at the Superchip

NVIDIA Grace Hopper Superchip

The NVIDIA GH200 Grace Hopper Superchip combines the Grace and Hopper architectures using NVIDIA® NVLink®-C2C to deliver a CPU+GPU coherent memory model for accelerated AI and HPC applications.

  • CPU+GPU designed for giant-scale AI and HPC
  • New 900 gigabytes per second (GB/s) coherent interface, 7X faster than PCIe Gen5
  • 30X higher aggregate system memory bandwidth to GPU compared to NVIDIA DGX™ A100
  • Runs all NVIDIA software stacks and platforms, including the NVIDIA HPC SDK, NVIDIA AI, and NVIDIA Omniverse™

Explore Grace Hopper Reference Design for Modern Data Center Workloads

HGX - AI Training, Inference & HPC

NVIDIA HGX™

for AI training, inference, and HPC.

NVIDIA GH200 Superchip
NVIDIA BlueField®-3
OEM-defined input/output (IO) and fourth-generation NVLink

GH200-powered systems join 400+ configurations—based on the latest NVIDIA architectures—that are being rolled out to meet the surging demand for generative AI.