Product Compare
NVIDIA GeForce RTX 3070 Ti GA102 vs NVIDIA H100 SXM5 80 GB
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Oct 2022 | Oct 2022 |
| Generation | GeForce 30 | Server Hopper |
| Type | Desktop | AI GPU |
| Bus Interface | PCIe 4.0 x16 | PCIe 5.0 x16 |
| Clock Speeds | ||
| Base Clock | 1575 MHz | 1590 MHz |
| Boost Clock | 1770 MHz | 1980 MHz |
| Memory Clock | 1188 MHz | 1313 MHz |
| Memory | ||
| Memory Size | 8GB | 80GB |
| Memory Type | GDDR6X | HBM3 |
| Memory Bus | 256bit | 5120bit |
| Bandwidth | 608.3GB/s | 3.36TB/s |
| Render Config | ||
| SM Count | 48 | 132 |
| Shading Units | 6144 | 16896 |
| TMUs | 192 | 528 |
| ROPs | 96 | 24 |
| Tensor Cores | 192 | 528 |
| RT Cores | 48 | |
| L1 Cache | 128 KB (per SM) | 256 KB (per SM) |
| L2 Cache | 4 MB | 50 MB |
| Theoretical Performance | ||
| Pixel Rate | 169.9 GPixel/s | 47.52 GPixel/s |
| Texture Rate | 339.8 GTexel/s | 1045 GTexel/s |
| FP16 (half) | 21.75 TFLOPS | 267.6 TFLOPS |
| FP32 (float) | 21.75 TFLOPS | 66.91 TFLOPS |
| FP64 (double) | 339.8 GFLOPS | 33.45 TFLOPS |
| Graphics Processor | ||
| GPU Name | GA102 | GH100 |
| GPU Variant | GA102-150-A1 | |
| Architecture | Ampere | Hopper |
| Foundry | Samsung | TSMC |
| Process Size | 8 nm | 5 nm |
| Transistors | 28.3 billion | 80 billion |
| Die Size | 628 mm² | 814 mm² |
| Board Design | ||
| TDP | 290W | 700W |
| Suggested PSU | 600 W | 1100 W |
| Outputs | 1x HDMI 2.13x DisplayPort 1.4a | No outputs |
| Power Connectors | 1x 12-pin | None |
| Graphics Features | ||
| DirectX | 12 Ultimate (12_2) | N/A |
| OpenGL | 4.6 | N/A |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.3 | N/A |
| CUDA | 8.6 | 9.0 |
| Shader Model | 6.6 | N/A |