NVIDIA H800 Enterprise 80GB
AI adoption has become mainstream in enterprises, driving the need for comprehensive, AI-ready infrastructure to propel organizations into this new era.
The H800 for mainstream servers includes a five-year subscription to the NVIDIA AI Enterprise software suite, featuring enterprise support to streamline AI integration with top-tier performance. This solution provides access to essential AI frameworks and tools, enabling organizations to develop H800-accelerated AI workflows, such as AI chatbots, recommendation engines, vision AI, and more.
Contact us for pricing
Tech Specs
Specification | Details |
|---|---|
Graphics Processor | |
GPU Name | GH100 |
Architecture | Hopper |
Foundry | TSMC |
Process Size | 4 nm |
Transistors | 80,000 million |
Die Size | 814mm² |
Graphics Card | |
Release Date | Mar 2022 |
Generation | Tesla Hopper |
Type | Desktop |
Bus Interface | PCIe 5.0 x16 |
Clock Speeds | |
Base Clock | 1095MHz |
Boost Clock | 1755MHz |
Memory Clock | 1500MHz |
Memory | |
Memory Size | 80GB |
Memory Type | HBM3 |
Memory Bus | 5120bit |
Bandwidth | 1920 GB/s |
Render Config | |
Shading Units | 8448 |
TMUs | 528 |
ROPs | 24 |
SM Count | 132 |
Tensor Cores | 528 |
L1 Cache | 192 KB (per SM) |
L2 Cache | 50MB |
Theoretical Performance | |
Pixel Rate | 42.12 GPixel/s |
Texture Rate | 926.6 GTexel/s |
FP16 (half) | 118.6 TFLOPS |
FP32 (float) | 29.65 TFLOPS |
FP64 (double) | 14.83 TFLOPS |
Board Design | |
TDP | 700W |
Suggested PSU | 1100W |
Outputs | No outputs |
Power Connectors | 8-pin EPS |
Graphics Features | |
DirectX | N/A |
OpenGL | N/A |
OpenCL | 3.0 |
Vulkan | N/A |
CUDA | 9.0 |
Shader Model | N/A |