NVIDIA Grace Hopper Quad GH200
The NVIDIA Quad GH200 Grace Hopper connects four GH200 superchips with 4th Generation NVLink to serve as the engine behind scientific computing.
NVIDIA Quad GH200
The NVIDIA GH200 Grace Hopper Superchip connects the high-performance and energy efficient Grace CPU with the H100 Tensor Core GPU with the high-bandwidth fully coherent 900 GB/s NVLink Chip-to-Chip (C2C). The NVIDIA Quad GH200 Grace Hopper connects four GH200 superchips with 4th Generation NVLink to serve as the engine behind generative AI and scientific computing, helping scientists and researchers reach unprecedented solutions for the world’s most complex problems.
Accelerated Computing and Generative AI
The NVIDIA GH200 Grace Hopper Superchip with a coherent link creates a unified memory address space to simplify model programming. It combines high-bandwidth and low-power system memory LPDDR5X and HBM to take full advantage of NVIDIA GPU acceleration and high-performance Arm cores supercharging performance for the latest accelerated computing and generative AI worklaods.
Versatile Platform
Enterprises need a versatile system to handle the largest models and realize the full potential of their infrastructure. The Quad GH200 is architected to deliver breakthrough performance for CPU intensive, memory intensive and compute intensive workloads, allowing customers to maximize their infrastructure.
Certifications
Red Hat certified hardware is proven to incorporate Red Hat's best practices and provides customers tested interoperability, known life cycle management, and trusted support.
| Compare | Product | Level | |
|---|---|---|---|
Red Hat Enterprise Linux 10.0 - 10.x Architecture: aarch64 |
