Loading the catalog…
H200 NVL (NVIDIA)
Blurred values are subscriber-only. Sign in to start a free trial — 2 full pages a day — or subscribe for unlimited access.
| Memory Bandwidth Tb S | 4.8 |
|---|---|
| Fp8 Tflops Dense | 1670.5 |
| Interconnect | 2- or 4-way NVIDIA NVLink bridge: 900GB/s per GPU |
| Fp32 Tflops | 60 |
| Tdp W | 600 |
| Memory Type | HBM3E |
| Cooling | air |
| Gpu Memory Gb | 141 |
| Accelerator Form Factor | PCIe |
| Architecture | Hopper |
| Fp16 Bf16 Tflops Dense | 835.5 |
| Fp16 Bf16 Tflops Sparse | 1671 |
| Int8 Tops Dense | 1670.5 |
| Fp64 Tflops | |
| Fp8 Tflops Sparse | |
| Int8 Tops Sparse | |
| Media Engines | |
| Product Name | |
| Tf32 Tflops Dense | |
| Tf32 Tflops Sparse |
| SKU | H200 NVL |
|---|
| BFLOAT16 Tensor Core | |
|---|---|
| INT8 Tensor Core | |
| Confidential Computing | |
| Multi-Instance GPUs | |
| Server Options | |
| NVIDIA AI Enterprise | |
| NVIDIA AI Enterprise subscription | |
| Max GPUs connected by NVLink | |
| Memory increase vs H100 NVL | |
| LLM inference acceleration vs H100 NVL | |
| HPC performance vs H100 NVL | |
| Footnote 1 | |
| Footnote 2 |
