Product Compare
AMD Radeon Instinct MI350X vs NVIDIA L20
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Jun 2025 | Nov 2023 |
| Generation | Radeon Instinct | Server Ada |
| Type | Professional | Desktop |
| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x16 |
| Clock Speeds | ||
| Base Clock | 1000 MHz | 1440 MHz |
| Boost Clock | 2200 MHz | 2520 MHz |
| Memory Clock | 2000 MHz | 2250 MHz |
| Memory | ||
| Memory Size | 288GB | 48GB |
| Memory Type | HBM3e | GDDR6 |
| Memory Bus | 8192bit | 384bit |
| Bandwidth | 8.19TB/s | 864.0GB/s |
| Render Config | ||
| Compute Units | 256 | |
| Shading Units | 16384 | 11776 |
| TMUs | 1024 | 368 |
| ROPs | 0 | 128 |
| Tensor Cores | 1024 | 368 |
| L1 Cache | 16 KB (per CU) | 128 KB (per SM) |
| L2 Cache | 16 MB | 96 MB |
| L3 Cache | 256 MB | |
| SM Count | 92 | |
| RT Cores | 92 | |
| Theoretical Performance | ||
| Pixel Rate | 0 MPixel/s | 322.6 GPixel/s |
| Texture Rate | 2252.8 GTexel/s | 927.4 GTexel/s |
| FP16 (half) | 72.09 TFLOPS | 59.35 TFLOPS |
| FP32 (float) | 72.09 TFLOPS | 59.35 TFLOPS |
| FP64 (double) | 36.04 TFLOPS | 927.4 GFLOPS |
| Graphics Processor | ||
| GPU Name | MI350 256CU | AD102 |
| Architecture | CDNA 4.0 | Ada Lovelace |
| Foundry | TSMC | TSMC |
| Process Size | 3 nm | 5 nm |
| Transistors | 185 billion | 76.3 billion |
| Die Size | Unknown | 609 mm² |
| GPU Variant | AD102 | |
| Board Design | ||
| TDP | 1000W | 275W |
| Suggested PSU | 1400 W | 600 W |
| Outputs | No outputs | 4x DisplayPort 1.4a |
| Power Connectors | None | 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 |
| Shader Model | N/A | 6.9 |
| CUDA | 8.9 | |