Quick/Comparisons
59 | Gaming | 35 |
59 | Workstation | 38 |
56 | AI/ML |
Compare with something else?
Compare Radeon RX 6950 with another →38 |
39 | Energy Efficiency | 44 |
58 | Quick Comparison Final Score | 36 |
| Navi 21 | Graphics Processor | Navi 33 |
| 5120 | Cores | 2048 |
| 320 / 128 | TMUs / ROPs | 128 / 64 |
| 2746.09 | Blender GPU | 1271.08 |
| 142 | Forza Horizon 5 | 81 |
| 193 | The Witcher 3 | 107 |
| 216 | Counter-Strike 2 | 146 |
| 151 | Far Cry 6 | 99 |
| 107 | Hogwarts Legacy | 66 |
| 184 | Call of Duty | 97 |
| 115 | Ghost of Tsushima | 68 |
| 135 | Cyberpunk 2077 | 72 |
| 219 | Tomb Raider | 121 |
| 162 | Average FPS | 95 |
| 195 | 1080p High | 120 |
| 162 | 1080p Ultra | 95 |
| 124 | 1440p Ultra | 68 |
| 72 | 4K Ultra | 37 |
| Low | Margin of Error | Low |
| 189857 | GB6 Compute Score | 86899 |
| 382.1 img/sec | Background Blur | 199.6 img/sec |
| 259 img/sec | Face Detection | 106.8 img/sec |
| 6.93 Gpixels/sec | Horizon Detection | 3.23 Gpixels/sec |
| 10.2 Gpixels/sec | Edge Detection | 4.3 Gpixels/sec |
| 8.39 Gpixels/sec | Gaussian Blur | 3.92 Gpixels/sec |
| 1.48 Gpixels/sec | Feature Matching | 1 Gpixels/sec |
| 824.3 Gpixels/sec | Stereo Matching | 292.4 Gpixels/sec |
| 26538.1 FPS | Particle Physics | 10709.1 FPS |
| OpenCL | API | OpenCL |
| 28074 | G3D Mark Score | 17295 |
| 1079 | G2D Mark | 1027 |
| 299 FPS | DirectX 11 | 168 FPS |
| 115 FPS | DirectX 12 | 67 FPS |
| 14148 Ops/s | GPU Compute | 8951 Ops/s |
| 33294 | GB6 ML Single Precision | 20184 |
| 36737 | GB6 ML Half Precision | 26434 |
| 24655 | GB6 ML Quantized | 16381 |
| 14145 | Image Classification (SP) | 9915 |
| 18828 | Image Segmentation (HP) | 10595 |
| 32309 | Image Super Resolution (Q) | 25345 |
| 49133 | Face Detection (HP) | 33847 |
| 166767 | Pose Estimation (Q) | 92252 |
| 4491 | Text Classification (SP) | 3842 |
| 6690 | Machine Translation (HP) | 4096 |
| 15446 | Object Detection (SP) | 7170 |
| 56831 | Depth Estimation (Q) | 43860 |
| 367328 | Style Transfer (SP) | 185627 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 296 GPixel/s | Pixel Fill Rate | 176 GPixel/s |
| 739 GTexel/s | Texture Fill Rate | 353 GTexel/s |
| 23.7 TFLOPS | FLOPS (FP32) | 22.6 TFLOPS |
| GDDR6 | Memory Type | GDDR6 |
| 16 GB | Memory Size | 16 GB |
| 2250 MHz | Memory Clock | 2250 MHz |
| 18000 Mbps | Effective Memory Speed | 18000 Mbps |
| 256-bit | Bus | 128-bit |
| No | ECC | No |
| 576 GB/s | Memory Bandwidth | 288 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 4.0 x8 |
| 335 W | TGP | 190 W |
| TSMC | Manufacturing | TSMC |
| 7 nm | Fabrication Process | 6 nm |
| 520 mm² | Die Size | 204 mm² |
| 26 billion | Transistor Count | 13 billion |
| 50 MTr/mm² | Transistor Density | 63.73 MTr/mm² |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 2.1 | OpenCL | 2.2 |
| — | CUDA | — |
| Yes | Ray Tracing | Yes |
| No | DLSS | No |
| 1.4a | DisplayPort | 2.1 |
| Amd | Vendor | Amd |
| Discrete | Build | Discrete |
| May 10, 2022 | Released | January 24, 2024 |
| Desktop | Case | Desktop |
| Gaming | Purpose | Gaming |
| High-end | Segment | Mid-range |
| RDNA 2.0 | Architecture | RDNA 3.0 |
| Navi 21 | GPU Codename | Navi 33 |
| -Intel Core i7 12700Kor above | Recommended CPU | -Intel Core i5 13600Kor above |
| 20872 | Steel Nomad Lite Score | 10251 |
| 21598 | Time Spy | 11136 |
| 89230 | Solar Bay | 44382 |
| 11260 | Port Royal | 5890 |
| 59884 | Fire Strike | 31713 |
| 40953 | Wild Life Extreme | 20435 |
| 163724 | Night Raid | 122651 |
| 1860 MHz | Base Clock | 1980 MHz |
| 2310 MHz | Boost Clock | 2755 MHz |
| 5120 | Shading Units | 2048 |
| 320 | Texture Mapping Units (TMUs) | 128 |
| 128 | Render Output Units (ROPs) | 64 |
| — | Compute Units (Pipelines) | — |
| — | Tensor Cores | — |
| 80 | Ray-tracing Cores | 32 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 4MB shared | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 4 IPC |