Product Compare
NVIDIA RTX A2000 Max-Q vs AMD Instinct MI300A
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Apr 2021 | Dec 2023 |
| Generation | Ampere-MW | Radeon Instinct |
| Type | Mobile | Professional |
| Bus Interface | PCIe 4.0 x16 | PCIe 5.0 x16 |
| Clock Speeds | ||
| Base Clock | 682 MHz | 1000 MHz |
| Boost Clock | 1207 MHz | 2100 MHz |
| Memory Clock | 1375 MHz | 1300 MHz |
| Memory | ||
| Memory Size | 4GB | 128GB |
| Memory Type | GDDR6 | HBM3 |
| Memory Bus | 128bit | 8192bit |
| Bandwidth | 176.0GB/s | 5.32TB/s |
| Render Config | ||
| SM Count | 20 | |
| Shading Units | 2560 | 14592 |
| TMUs | 80 | 912 |
| ROPs | 48 | 0 |
| Tensor Cores | 80 | |
| RT Cores | 20 | |
| L1 Cache | 128 KB (per SM) | 16 KB (per CU) |
| L2 Cache | 2 MB | 16 MB |
| Compute Units | 228 | |
| L3 Cache | 256 MB | |
| Theoretical Performance | ||
| Pixel Rate | 57.94 GPixel/s | 0 MPixel/s |
| Texture Rate | 96.56 GTexel/s | 1915.2 GTexel/s |
| FP16 (half) | 6.180 TFLOPS | 653.7 TFLOPS |
| FP32 (float) | 6.180 TFLOPS | 61.29 TFLOPS |
| FP64 (double) | 96.56 GFLOPS | 30.64 TFLOPS |
| Graphics Processor | ||
| GPU Name | GA107 | Aqua Vanjaram |
| Architecture | Ampere | CDNA 3.0 |
| Foundry | Samsung | TSMC |
| Process Size | 8 nm | 5 nm |
| Transistors | 8.7 billion | 153 billion |
| Die Size | 200 mm² | 1017 mm² |
| Board Design | ||
| TDP | 95W | 750W |
| Outputs | Portable Device Dependent | No outputs |
| Power Connectors | None | None |
| Suggested PSU | 1150 W | |
| Graphics Features | ||
| DirectX | 12 Ultimate (12_2) | N/A |
| OpenGL | 4.6 | N/A |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.3 | N/A |
| CUDA | 8.6 | |
| Shader Model | 6.8 | N/A |