Product Compare
NVIDIA Rubin GPU vs NVIDIA L20
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Jan 2026 | Nov 2023 |
| Generation | Server Rubin | Server Ada |
| Type | Desktop | Desktop |
| Bus Interface | PCIe 6.0 x16 | PCIe 4.0 x16 |
| Clock Speeds | ||
| Base Clock | 700 MHz | 1440 MHz |
| Boost Clock | 2267 MHz | 2520 MHz |
| Memory Clock | 2695 MHz | 2250 MHz |
| Memory | ||
| Memory Size | 288GB | 48GB |
| Memory Type | HBM4 | GDDR6 |
| Memory Bus | 16384bit | 384bit |
| Bandwidth | 22.1TB/s | 864.0GB/s |
| Render Config | ||
| SM Count | 224 | 92 |
| Shading Units | 28672 | 11776 |
| TMUs | 896 | 368 |
| ROPs | 24 | 128 |
| Tensor Cores | 896 | 368 |
| L1 Cache | 256 KB (per SM) | 128 KB (per SM) |
| L2 Cache | 128 MB | 96 MB |
| RT Cores | 92 | |
| Theoretical Performance | ||
| Pixel Rate | 54.41 GPixel/s | 322.6 GPixel/s |
| Texture Rate | 2031.2 GTexel/s | 927.4 GTexel/s |
| FP16 (half) | 260.0 TFLOPS | 59.35 TFLOPS |
| FP32 (float) | 130.0 TFLOPS | 59.35 TFLOPS |
| FP64 (double) | 32.50 TFLOPS | 927.4 GFLOPS |
| Graphics Processor | ||
| GPU Name | GR100 | AD102 |
| Architecture | Rubin | Ada Lovelace |
| Foundry | TSMC | TSMC |
| Process Size | 3 nm | 5 nm |
| Transistors | 336 billion | 76.3 billion |
| Die Size | 1456 mm² | 609 mm² |
| GPU Variant | AD102 | |
| Board Design | ||
| TDP | 2300W | 275W |
| Suggested PSU | 2700 W | 600 W |
| Outputs | No outputs | 4x DisplayPort 1.4a |
| Power Connectors | 1x 16-pin | |
| Graphics Features | ||
| DirectX | N/A | 12 Ultimate (12_2) |
| OpenGL | N/A | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | N/A | 1.4 |
| CUDA | 10.7 | 8.9 |
| Shader Model | N/A | 6.9 |