Product Compare
NVIDIA RTX PRO 4500 Blackwell Server vs NVIDIA L40 CNX
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Mar 2026 | Oct 2022 |
| Generation | Server Blackwell | Tesla Ada |
| Type | Professional | Professional |
| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x16 |
| Clock Speeds | ||
| Base Clock | 1215 MHz | 1005 MHz |
| Boost Clock | 2415 MHz | 2475 MHz |
| Memory Clock | 1563 MHz | 2250 MHz |
| Memory | ||
| Memory Size | 32GB | 24GB |
| Memory Type | GDDR7 | GDDR6 |
| Memory Bus | 256bit | 384bit |
| Bandwidth | 800.3GB/s | 864.0GB/s |
| Render Config | ||
| SM Count | 82 | 142 |
| Shading Units | 10496 | 18176 |
| TMUs | 328 | 568 |
| ROPs | 112 | 192 |
| Tensor Cores | 328 | 568 |
| RT Cores | 82 | 142 |
| L1 Cache | 128 KB (per SM) | 128 KB (per SM) |
| L2 Cache | 64 MB | 48 MB |
| Theoretical Performance | ||
| Pixel Rate | 270.5 GPixel/s | 475.2 GPixel/s |
| Texture Rate | 792.1 GTexel/s | 1406 GTexel/s |
| FP16 (half) | 50.70 TFLOPS | 89.97 TFLOPS |
| FP32 (float) | 50.70 TFLOPS | 89.97 TFLOPS |
| FP64 (double) | 792.1 GFLOPS | 1406 GFLOPS |
| Graphics Processor | ||
| GPU Name | GB203 | AD102 |
| Architecture | Blackwell 2.0 | Ada Lovelace |
| Foundry | TSMC | TSMC |
| Process Size | 5 nm | 4 nm |
| Transistors | 45.6 billion | 76.3 billion |
| Die Size | 378 mm² | 608 mm² |
| GPU Variant | AD102-???-A1 | |
| Board Design | ||
| TDP | 165W | 300W |
| Suggested PSU | 450 W | 700 W |
| Outputs | No outputs | 1x HDMI 2.13x DisplayPort 1.4a |
| Power Connectors | 1x 16-pin | 1x 16-pin |
| Graphics Features | ||
| DirectX | 12 Ultimate (12_2) | 12 Ultimate (12_2) |
| OpenGL | 4.6 | 4.6 |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.4 | 1.3 |
| CUDA | 12.0 | 8.9 |
| Shader Model | 6.9 | 6.6 |