Quick/Comparisons
The Nvidia RTX Pro 6000 Blackwell's exceptional gaming performance, with a score of 99, leaves the Radeon RX 6950 XT far behind with a score of 59.
The Nvidia RTX Pro 6000 Blackwell's perfect score of 100 in workstation performance makes it the clear winner, while the Radeon RX 6950 XT trails behind with a score of 59.
Although the Nvidia RTX Pro 6000 Blackwell's AI/ML score of 93 is not a perfect 100, it still outperforms the Radeon RX 6950 XT's score of 56 by a significant margin.
99 | Gaming | 59 |
100 | Workstation | 59 |
93 | AI/ML |
Compare with something else?
Compare Nvidia RTX Pro with another →56 |
49 | Energy Efficiency | 39 |
98 | Quick Comparison Final Score | 58 |
| GB202-300-A1 | Graphics Processor | Navi 21 |
| 24064 | Cores | 5120 |
| 752 / 190 | TMUs / ROPs | 320 / 128 |
| 16644.8 | Blender GPU | 2746.09 |
| 230 | Forza Horizon 5 | 142 |
| 304 | The Witcher 3 | 193 |
| 259 | Counter-Strike 2 | 216 |
| 198 | Far Cry 6 | 151 |
| 192 | Hogwarts Legacy | 107 |
| 301 | Call of Duty | 184 |
| 172 | Ghost of Tsushima | 115 |
| 221 | Cyberpunk 2077 | 135 |
| 303 | Tomb Raider | 219 |
| 242 | Average FPS | 162 |
| 282 | 1080p High | 195 |
| 242 | 1080p Ultra | 162 |
| 197 | 1440p Ultra | 124 |
| 118 | 4K Ultra | 72 |
| High | Margin of Error | Low |
| 419470 | GB6 Compute Score | 189857 |
| 366.1 img/sec | Background Blur | 382.1 img/sec |
| 248 img/sec | Face Detection | 259 img/sec |
| 23.3 Gpixels/sec | Horizon Detection | 6.93 Gpixels/sec |
| 35.2 Gpixels/sec | Edge Detection | 10.2 Gpixels/sec |
| 45 Gpixels/sec | Gaussian Blur | 8.39 Gpixels/sec |
| 2.47 Gpixels/sec | Feature Matching | 1.48 Gpixels/sec |
| 3040 Gpixels/sec | Stereo Matching | 824.3 Gpixels/sec |
| 42574.6 FPS | Particle Physics | 26538.1 FPS |
| OpenCL | API | OpenCL |
| 37305 | G3D Mark Score | 28074 |
| 1375 | G2D Mark | 1079 |
| 338 FPS | DirectX 11 | 299 FPS |
| 153 FPS | DirectX 12 | 115 FPS |
| 26895 Ops/s | GPU Compute | 14148 Ops/s |
| 50053 | GB6 ML Single Precision | 33294 |
| 74036 | GB6 ML Half Precision | 36737 |
| 37636 | GB6 ML Quantized | 24655 |
| 16903 | Image Classification (SP) | 14145 |
| 48518 | Image Segmentation (HP) | 18828 |
| 54618 | Image Super Resolution (Q) | 32309 |
| 103224 | Face Detection (HP) | 49133 |
| 338587 | Pose Estimation (Q) | 166767 |
| 3894 | Text Classification (SP) | 4491 |
| 7291 | Machine Translation (HP) | 6690 |
| 26498 | Object Detection (SP) | 15446 |
| 90853 | Depth Estimation (Q) | 56831 |
| 635289 | Style Transfer (SP) | 367328 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 497 GPixel/s | Pixel Fill Rate | 296 GPixel/s |
| 1968 GTexel/s | Texture Fill Rate | 739 GTexel/s |
| 126 TFLOPS | FLOPS (FP32) | 23.7 TFLOPS |
| GDDR7 | Memory Type | GDDR6 |
| 96 GB | Memory Size | 16 GB |
| 1750 MHz | Memory Clock | 2250 MHz |
| 28000 Mbps | Effective Memory Speed | 18000 Mbps |
| 512-bit | Bus | 256-bit |
| Yes | ECC | No |
| 1792 GB/s | Memory Bandwidth | 576 GB/s |
| PCIe 5.0 x16 | Interface | PCIe 4.0 x16 |
| 600 W | TGP | 335 W |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 7 nm |
| 750 mm² | Die Size | 520 mm² |
| 92.2 billion | Transistor Count | 26 billion |
| 122.93 MTr/mm² | Transistor Density | 50 MTr/mm² |
| 12.2 | DirectX | 12 |
| 1.4 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 2.1 |
| 12 | CUDA | — |
| Yes | Ray Tracing | Yes |
| DLSS 4 | DLSS | No |
| 2.1b | DisplayPort | 1.4a |
| Nvidia | Vendor | Amd |
| Discrete | Build | Discrete |
| March 18, 2026 | Released | May 10, 2022 |
| Desktop | Case | Desktop |
| Professional | Purpose | Gaming |
| High-end | Segment | High-end |
| Blackwell 2.0 | Architecture | RDNA 2.0 |
| GB202-300-A1 | GPU Codename | Navi 21 |
| -AMD Ryzen 9 9955HX3Dor above | Recommended CPU | -Intel Core i7 12700Kor above |
| 50516 | Steel Nomad Lite Score | 20872 |
| 42437 | Time Spy | 21598 |
| 258808 | Solar Bay | 89230 |
| 39952 | Port Royal | 11260 |
| 72183 | Fire Strike | 59884 |
| 110070 | Wild Life Extreme | 40953 |
| 216213 | Night Raid | 163724 |
| 1590 MHz | Base Clock | 1860 MHz |
| 2617 MHz | Boost Clock | 2310 MHz |
| 24064 | Shading Units | 5120 |
| 752 | Texture Mapping Units (TMUs) | 320 |
| 190 | Render Output Units (ROPs) | 128 |
| 188 | Compute Units (Pipelines) | — |
| 752 | Tensor Cores | — |
| 188 | Ray-tracing Cores | 80 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 128MB shared | L2 Cache | 4MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |