Product Compare
Intel Data Center GPU Max Subsystem vs NVIDIA H100 SXM5 80 GB
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Jan 2023 | Oct 2022 |
| Generation | Data Center GPU | Server Hopper |
| Type | Professional | AI GPU |
| Bus Interface | PCIe 5.0 x16 | PCIe 5.0 x16 |
| Clock Speeds | ||
| Base Clock | 900 MHz | 1590 MHz |
| Boost Clock | 1600 MHz | 1980 MHz |
| Memory Clock | 1565 MHz | 1313 MHz |
| Memory | ||
| Memory Size | 128GB | 80GB |
| Memory Type | HBM2e | HBM3 |
| Memory Bus | 8192bit | 5120bit |
| Bandwidth | 3205GB/s | 3.36TB/s |
| Render Config | ||
| Shading Units | 16384 | 16896 |
| TMUs | 1024 | 528 |
| ROPs | 0 | 24 |
| Tensor Cores | 1024 | 528 |
| RT Cores | 128 | |
| L1 Cache | 64 KB (per EU) | 256 KB (per SM) |
| L2 Cache | 408 MB | 50 MB |
| SM Count | 132 | |
| Theoretical Performance | ||
| Pixel Rate | 0 MPixel/s | 47.52 GPixel/s |
| Texture Rate | 1638 GTexel/s | 1045 GTexel/s |
| FP16 (half) | 52.43 TFLOPS | 267.6 TFLOPS |
| FP32 (float) | 52.43 TFLOPS | 66.91 TFLOPS |
| FP64 (double) | 52.43 TFLOPS | 33.45 TFLOPS |
| Graphics Processor | ||
| GPU Name | Ponte Vecchio | GH100 |
| Architecture | Generation 12.5 | Hopper |
| Foundry | Intel | TSMC |
| Process Size | 10 nm | 5 nm |
| Transistors | 100 billion | 80 billion |
| Die Size | 1280 mm² | 814 mm² |
| Board Design | ||
| TDP | 2400W | 700W |
| Suggested PSU | 2800 W | 1100 W |
| Outputs | No outputs | No outputs |
| Power Connectors | 1x 16-pin | None |
| Graphics Features | ||
| DirectX | 12 (12_1) | N/A |
| OpenGL | 4.6 | N/A |
| OpenCL | 3.0 | 3.0 |
| Vulkan | N/A | N/A |
| Shader Model | 6.6 | N/A |
| CUDA | 9.0 | |