NVIDIA CMP 40HX vs AMD Radeon Pro Vega 48
What is the difference between NVIDIA CMP 40HX and AMD Radeon Pro Vega 48. Find out which graphics card has better performance.
Graphics Processor (GPU)
| TU106 | GPU Name | Vega 10 |
| TU106-100-A1 | Model number | Vega 10 PRO |
| Turing | Architecture | GCN 5.0 |
| TSMC | Foundry | GlobalFoundries |
| 12 nm | Process Size | 14 nm |
| 10,800 million | Transistors | 12,500 million |
| 445 mm² | Die Size | 495 mm² |
Graphics Card
| Feb 25th, 2021 | Release Date | Mar 19th, 2019 |
| Mining GPUs | Family | Radeon Pro Mac (Vega Series) |
| Active | Production | Active |
| PCIe 3.0 x16 | Bus Interface | PCIe 3.0 x16 |
Memory
| 8 GB | Memory Size | 8 GB |
| GDDR6 | Memory Type | HBM2 |
| 256 bit | Memory Bus | 2048 bit |
| 448.0 GB/s | Bandwidth | 402.4 GB/s |
Performance
| 105.6 GPixel/s | Pixel fillrate | 76.80 GPixel/s |
| 237.6 GTexel/s | Texture fillrate | 230.4 GTexel/s |
| 15.21 TFLOPS (2:1) | FP16 (half) performance | 14.75 TFLOPS (2:1) |
| 7.603 TFLOPS | FP32 (float) performance | 7.373 TFLOPS |
| 237.6 GFLOPS (1:32) | FP64 (double) performance | 460.8 GFLOPS (1:16) |
Clock Speeds
| 1750 MHz 14 Gbps effective | Memory Clock | 786 MHz 1572 Mbps effective |
Render Config
| 2304 | Shading Units | 3072 |
| 144 | TMUs | 192 |
| 64 | ROPs | 64 |
| 64 KB (per SM) | L1 Cache | 16 KB (per CU) |
| 4 MB | L2 Cache | 4 MB |
Board Design
| Dual-slot | Slot Width | IGP |
| 185 W | Thermal design power (TDP) | unknown |
| No outputs | Display Connectors | No outputs |
| 1x 8-pin | Power Connectors | None |
API support
| 12 Ultimate (12_2) | DirectX | 12 (12_1) |
| 4.6 | OpenGL | 4.6 |
| 3.0 | OpenCL | 2.1 |
| 1.2 | Vulkan | 1.2 |
| 6.6 | Shader Model | 6.4 |