Product Compare
NVIDIA H100 CNX vs NVIDIA Tesla K20X
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Mar 2023 | Nov 2012 |
| Generation | Tesla Hopper | Tesla |
| Type | AI GPU | Professional |
| Bus Interface | PCIe 5.0 x16 | PCIe 3.0 x16 |
| Clock Speeds | ||
| Base Clock | 690 MHz | |
| Boost Clock | 1845 MHz | |
| Memory Clock | 1593 MHz | 1300 MHz |
| Memory | ||
| Memory Size | 80GB | 6GB |
| Memory Type | HBM2e | GDDR5 |
| Memory Bus | 5120bit | 384bit |
| Bandwidth | 2039GB/s | 249.6GB/s |
| Render Config | ||
| SM Count | 114 | |
| Shading Units | 14592 | 2688 |
| TMUs | 456 | 224 |
| ROPs | 24 | 48 |
| Tensor Cores | 456 | |
| L1 Cache | 256 KB (per SM) | 16 KB (per SMX) |
| L2 Cache | 50 MB | 1536 KB |
| Theoretical Performance | ||
| Pixel Rate | 44.28 GPixel/s | 40.99 GPixel/s |
| Texture Rate | 841.3 GTexel/s | 164.0 GTexel/s |
| FP16 (half) | 215.4 TFLOPS | |
| FP32 (float) | 53.84 TFLOPS | 3.935 TFLOPS |
| FP64 (double) | 26.92 TFLOPS | 1312 GFLOPS |
| Graphics Processor | ||
| GPU Name | GH100 | GK110 |
| Architecture | Hopper | Kepler |
| Foundry | TSMC | TSMC |
| Process Size | 4 nm | 28 nm |
| Transistors | 80 billion | 7.08 billion |
| Die Size | 814 mm² | 561 mm² |
| Board Design | ||
| TDP | 350W | 235W |
| Suggested PSU | 750 W | 550 W |
| Outputs | No outputs | No outputs |
| Power Connectors | 8-pin EPS | 1x 6-pin + 1x 8-pin |
| Graphics Features | ||
| DirectX | N/A | 12 (11_0) |
| OpenGL | N/A | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | N/A | 1.1 |
| CUDA | 9.0 | 3.5 |
| Shader Model | N/A | 5.1 |