Product Compare
AMD Radeon Instinct MI300A vs NVIDIA GeForce RTX 3070
| GPU | GPU | |
|---|---|---|
| Graphics Card | ||
| Release Date | Dec 2023 | Sep 2020 |
| Generation | Radeon Instinct | GeForce 30 |
| Type | Desktop | Desktop |
| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x16 |
| Clock Speeds | ||
| Base Clock | 1000 MHz | 1500 MHz |
| Boost Clock | 2100 MHz | 1725 MHz |
| Memory Clock | 2525 MHz | 1750 MHz |
| Memory | ||
| Memory Size | 192GB | 8GB |
| Memory Type | HBM3 | GDDR6 |
| Memory Bus | 8192bit | 256bit |
| Bandwidth | 10.3TB/s | 448.0GB/s |
| Render Config | ||
| Compute Units | 304 | |
| Shading Units | 19456 | 5888 |
| TMUs | 1216 | 184 |
| ROPs | 0 | 96 |
| Tensor Cores | 1216 | 184 |
| L1 Cache | 16 KB (per CU) | 128 KB (per SM) |
| L2 Cache | 16 MB | 4 MB |
| L3 Cache | 256 MB | |
| SM Count | 46 | |
| RT Cores | 46 | |
| Theoretical Performance | ||
| Pixel Rate | 0 MPixel/s | 165.6 GPixel/s |
| Texture Rate | 2554 GTexel/s | 317.4 GTexel/s |
| FP16 (half) | 653.7 TFLOPS | 20.31 TFLOPS |
| FP32 (float) | 81.72 TFLOPS | 20.31 TFLOPS |
| FP64 (double) | 81.72 TFLOPS | 317.4 GFLOPS |
| Graphics Processor | ||
| GPU Name | Aqua Vanjaram | GA104 |
| Architecture | CDNA 3.0 | Ampere |
| Foundry | TSMC | Samsung |
| Process Size | 5 nm | 8 nm |
| Transistors | 153 billion | 17.4 billion |
| Die Size | 1017 mm² | 392 mm² |
| GPU Variant | GA104-300-A1 | |
| Board Design | ||
| TDP | 750W | 220W |
| Suggested PSU | 1150 W | 550 W |
| Outputs | No outputs | 1x HDMI 2.13x DisplayPort 1.4a |
| Power Connectors | None | 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 | 6.6 |
| CUDA | 8.6 | |