AMD Radeon Instinct MI100 vs NVIDIA A100 PCIe
What is the difference between AMD Radeon Instinct MI100 and NVIDIA A100 PCIe. Find out which graphics card has better performance.
Graphics Processor (GPU)
| Arcturus | GPU Name | GA100 |
| CDNA 1.0 | Architecture | Ampere |
| TSMC | Foundry | TSMC |
| 7 nm | Process Size | 7 nm |
| 50,000 million | Transistors | 54,200 million |
| 750 mm² | Die Size | 826 mm² |
Graphics Card
| Nov 16th, 2020 | Release Date | Jun 22nd, 2020 |
| Radeon Instinct (MIx) | Family | Tesla (Axx) |
| Active | Production | Active |
| PCIe 4.0 x16 | Bus Interface | PCIe 4.0 x16 |
Memory
| 32 GB | Memory Size | 40 GB |
| HBM2 | Memory Type | HBM2e |
| 4096 bit | Memory Bus | 5120 bit |
| 1,229 GB/s | Bandwidth | 1,555 GB/s |
Performance
| 96.13 GPixel/s | Pixel fillrate | 225.6 GPixel/s |
| 721.0 GTexel/s | Texture fillrate | 609.1 GTexel/s |
| 184.6 TFLOPS (8:1) | FP16 (half) performance | 77.97 TFLOPS (4:1) |
| 23.07 TFLOPS | FP32 (float) performance | 19.49 TFLOPS |
| 11.54 TFLOPS (1:2) | FP64 (double) performance | 9.746 TFLOPS (1:2) |
Clock Speeds
| 1000 MHz | Base Clock | 765 MHz |
| 1502 MHz | Boost Clock | 1410 MHz |
| 1200 MHz 2.4 Gbps effective | Memory Clock | 1215 MHz 2.4 Gbps effective |
Render Config
| 7680 | Shading Units | 6912 |
| 480 | TMUs | 432 |
| 64 | ROPs | 160 |
| 16 KB (per CU) | L1 Cache | 192 KB (per SM) |
| 8 MB | L2 Cache | 40 MB |
Board Design
| Dual-slot | Slot Width | Dual-slot |
| 267 mm 10.5 inches | Length | 267 mm 10.5 inches |
| 300 W | Thermal design power (TDP) | 250 W |
| 700 W | Suggested PSU | 600 W |
| No outputs | Display Connectors | No outputs |
| 2x 8-pin | Power Connectors | 8-pin EPS |
| 109-D05187 | Board Number | P1001 SKU 200 |
API support
| N/A | DirectX | N/A |
| N/A | OpenGL | N/A |
| 2.1 | OpenCL | 3.0 |
| N/A | Vulkan | N/A |
| N/A | Shader Model | N/A |