Product Compare
NVIDIA GeForce RTX 5090 D vs AMD Radeon Instinct MI100
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Jan 2025 | Nov 2020 |
| Generation | GeForce 50 | Radeon Instinct |
| Type | Desktop | Professional |
| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x16 |
| Clock Speeds | ||
| Base Clock | 2017 MHz | 1000 MHz |
| Boost Clock | 2407 MHz | 1502 MHz |
| Memory Clock | 1750 MHz | 1200 MHz |
| Memory | ||
| Memory Size | 32GB | 32GB |
| Memory Type | GDDR7 | HBM2 |
| Memory Bus | 512bit | 4096bit |
| Bandwidth | 1.79TB/s | 1229GB/s |
| Render Config | ||
| SM Count | 170 | |
| Shading Units | 21760 | 7680 |
| TMUs | 680 | 480 |
| ROPs | 176 | 64 |
| Tensor Cores | 680 | |
| RT Cores | 170 | |
| L1 Cache | 128 KB (per SM) | 16 KB (per CU) |
| L2 Cache | 96 MB | 8 MB |
| Compute Units | 120 | |
| Theoretical Performance | ||
| Pixel Rate | 423.6 GPixel/s | 96.13 GPixel/s |
| Texture Rate | 1636.8 GTexel/s | 721.0 GTexel/s |
| FP16 (half) | 104.8 TFLOPS | 184.6 TFLOPS |
| FP32 (float) | 104.8 TFLOPS | 23.07 TFLOPS |
| FP64 (double) | 1.637 TFLOPS | 11.54 TFLOPS |
| Graphics Processor | ||
| GPU Name | GB202 | Arcturus |
| GPU Variant | GB202-250-A1 | Arcturus XL |
| Architecture | Blackwell 2.0 | CDNA 1.0 |
| Foundry | TSMC | TSMC |
| Process Size | 5 nm | 7 nm |
| Transistors | 92.2 billion | 25.6 billion |
| Die Size | 750 mm² | 750 mm² |
| Board Design | ||
| TDP | 575W | 300W |
| Suggested PSU | 950 W | 700 W |
| Outputs | 1x HDMI 2.1b3x DisplayPort 2.1b | No outputs |
| Power Connectors | 1x 16-pin | 2x 8-pin |
| Graphics Features | ||
| DirectX | 12 Ultimate (12_2) | N/A |
| OpenGL | 4.6 | N/A |
| OpenCL | 3.0 | 2.1 |
| Vulkan | 1.4 | N/A |
| CUDA | 12.0 | |
| Shader Model | 6.8 | N/A |