Product Compare
NVIDIA Rubin GPU vs NVIDIA GeForce RTX 3090
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Jan 2026 | Sep 2020 |
| Generation | Server Rubin | GeForce 30 |
| Type | Desktop | Desktop |
| Bus Interface | PCIe 6.0 x16 | PCIe 4.0 x16 |
| Clock Speeds | ||
| Base Clock | 700 MHz | 1395 MHz |
| Boost Clock | 2267 MHz | 1695 MHz |
| Memory Clock | 2695 MHz | 1219 MHz |
| Memory | ||
| Memory Size | 288GB | 24GB |
| Memory Type | HBM4 | GDDR6X |
| Memory Bus | 16384bit | 384bit |
| Bandwidth | 22.1TB/s | 936.2GB/s |
| Render Config | ||
| SM Count | 224 | 82 |
| Shading Units | 28672 | 10496 |
| TMUs | 896 | 328 |
| ROPs | 24 | 112 |
| Tensor Cores | 896 | 328 |
| L1 Cache | 256 KB (per SM) | 128 KB (per SM) |
| L2 Cache | 128 MB | 6 MB |
| RT Cores | 82 | |
| Theoretical Performance | ||
| Pixel Rate | 54.41 GPixel/s | 189.8 GPixel/s |
| Texture Rate | 2031.2 GTexel/s | 556.0 GTexel/s |
| FP16 (half) | 260.0 TFLOPS | 35.58 TFLOPS |
| FP32 (float) | 130.0 TFLOPS | 35.58 TFLOPS |
| FP64 (double) | 32.50 TFLOPS | 556.0 GFLOPS |
| Graphics Processor | ||
| GPU Name | GR100 | GA102 |
| Architecture | Rubin | Ampere |
| Foundry | TSMC | Samsung |
| Process Size | 3 nm | 8 nm |
| Transistors | 336 billion | 28.3 billion |
| Die Size | 1456 mm² | 628 mm² |
| GPU Variant | GA102-300-A1 | |
| Board Design | ||
| TDP | 2300W | 350W |
| Suggested PSU | 2700 W | 750 W |
| Outputs | No outputs | 1x HDMI 2.13x DisplayPort 1.4a |
| Power Connectors | 1x 12-pin | |
| Graphics Features | ||
| DirectX | N/A | 12 Ultimate (12_2) |
| OpenGL | N/A | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | N/A | 1.3 |
| CUDA | 10.7 | 8.6 |
| Shader Model | N/A | 6.6 |