Product Compare
AMD Radeon Instinct MI250X vs NVIDIA CMP 70HX
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Nov 2021 | Jan 2021 |
| Generation | Radeon Instinct | Mining GPUs |
| Type | Professional | Desktop |
| Bus Interface | PCIe 4.0 x16 | PCIe 1.0 x4 |
| Clock Speeds | ||
| Base Clock | 1000 MHz | 1365 MHz |
| Boost Clock | 1700 MHz | 1395 MHz |
| Memory Clock | 1600 MHz | 1188 MHz |
| Memory | ||
| Memory Size | 128GB | 8GB |
| Memory Type | HBM2e | GDDR6X |
| Memory Bus | 8192bit | 256bit |
| Bandwidth | 3277GB/s | 608.3GB/s |
| Render Config | ||
| Compute Units | 220 | |
| Shading Units | 14080 | 3840 |
| TMUs | 880 | 120 |
| ROPs | 0 | 64 |
| L1 Cache | 16 KB (per CU) | 128 KB (per SM) |
| L2 Cache | 16 MB | 4 MB |
| SM Count | 30 | |
| Tensor Cores | 120 | |
| RT Cores | 30 | |
| Theoretical Performance | ||
| Pixel Rate | 0 MPixel/s | 89.28 GPixel/s |
| Texture Rate | 1496 GTexel/s | 167.4 GTexel/s |
| FP16 (half) | 383.0 TFLOPS | 10.71 TFLOPS |
| FP32 (float) | 47.87 TFLOPS | 10.71 TFLOPS |
| FP64 (double) | 47.87 TFLOPS | 167.4 GFLOPS |
| Graphics Processor | ||
| GPU Name | Aldebaran | GA104 |
| GPU Variant | Aldebaran XT | GA104-100-A1 |
| Architecture | CDNA 2.0 | Ampere |
| Foundry | TSMC | Samsung |
| Process Size | 6 nm | 8 nm |
| Transistors | 58.2 billion | 17.4 billion |
| Die Size | Unknown | 392 mm² |
| Board Design | ||
| TDP | 500W | Unknown |
| Suggested PSU | 900 W | 200 W |
| Outputs | No outputs | No outputs |
| Power Connectors | 2x 8-pin | 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 |
| Shader Model | N/A | 6.8 |
| CUDA | 8.6 | |