Key Applications
High Performance Computing, AI/Deep Learning Training and Inference, Large
Language Model (LLM) and Generative AI,
Key Features
High density 1U GPU system with Integrated NVIDIA® H100 GPU;
NVIDIA Grace Hopper™ Superchip (Grace CPU and H100 GPU);
NVLink® Chip-2-Chip (C2C) high-bandwidth, low-latency interconnect
between CPU and GPU at 900GB/s;
Up to 576GB of coherent memory per node including 480GB LPDDR5X and
96GB of HBM3 for LLM applications;
2x PCIe 5.0 x16 slots supporting NVIDIA BlueField®-3 or ConnectX®-7;
9 Hot-Swap Heavy Duty Fans with Optimal Fan Speed Control;
This system supports two E1.S drives directly from the processor only.;