Quick/Comparisons
The Radeon RX 7900 GRE's superior gaming performance and higher score of 64 make it the clear winner in this use case. Its 26-point lead over the GeForce RTX 3060 Ti is substantial.
Although the Radeon RX 7900 GRE has a 18-point lead, the GeForce RTX 3060 Ti's workstation score of 42 is still respectable, but not enough to claim victory. The Radeon RX 7900 GRE's higher score gives it a slight edge.
The Radeon RX 7900 GRE's score of 59 in AI/ML is significantly higher than the GeForce RTX 3060 Ti's score of 31, making it the clear winner in this use case.
38 | Gaming | 64 |
42 | Workstation | 60 |
31 | AI/ML |
Compare with something else?
Compare GeForce RTX 3060 with another →59 |
35 | Energy Efficiency | 44 |
38 | Quick Comparison Final Score | 62 |
| GA104 | Graphics Processor | Navi 31 |
| 4864 | Cores | 5120 |
| 152 / 80 | TMUs / ROPs | 320 / 160 |
| 2992.04 | Blender GPU | 2808.96 |
| 87 | Forza Horizon 5 | 143 |
| 124 | The Witcher 3 | 201 |
| 167 | Counter-Strike 2 | 217 |
| 106 | Far Cry 6 | 155 |
| 71 | Hogwarts Legacy | 109 |
| 100 | Call of Duty | 190 |
| 68 | Ghost of Tsushima | 116 |
| 82 | Cyberpunk 2077 | 138 |
| 128 | Tomb Raider | 221 |
| 104 | Average FPS | 166 |
| 127 | 1080p High | 198 |
| 104 | 1080p Ultra | 166 |
| 77 | 1440p Ultra | 128 |
| 43 | 4K Ultra | 74 |
| Low | Margin of Error | Low |
| 115518 | GB6 Compute Score | 172224 |
| 185.1 img/sec | Background Blur | 352.4 img/sec |
| 108.4 img/sec | Face Detection | 229 img/sec |
| 5.35 Gpixels/sec | Horizon Detection | 6.42 Gpixels/sec |
| 7.17 Gpixels/sec | Edge Detection | 8.86 Gpixels/sec |
| 5.54 Gpixels/sec | Gaussian Blur | 7.89 Gpixels/sec |
| 1.15 Gpixels/sec | Feature Matching | 1.5 Gpixels/sec |
| 478 Gpixels/sec | Stereo Matching | 662.8 Gpixels/sec |
| 15183 FPS | Particle Physics | 24012 FPS |
| OpenCL | API | OpenCL |
| 20272 | G3D Mark Score | 27367 |
| 990 | G2D Mark | 1207 |
| 164 FPS | DirectX 11 | 300 FPS |
| 78 FPS | DirectX 12 | 110 FPS |
| 9905 Ops/s | GPU Compute | 15171 Ops/s |
| 17811 | GB6 ML Single Precision | 34533 |
| 29969 | GB6 ML Half Precision | 42224 |
| 11367 | GB6 ML Quantized | 26568 |
| 7384 | Image Classification (SP) | 15095 |
| 19320 | Image Segmentation (HP) | 21206 |
| 16108 | Image Super Resolution (Q) | 33962 |
| 34615 | Face Detection (HP) | 57614 |
| 74811 | Pose Estimation (Q) | 214748 |
| 2638 | Text Classification (SP) | 4325 |
| 4136 | Machine Translation (HP) | 6598 |
| 8718 | Object Detection (SP) | 16608 |
| 25946 | Depth Estimation (Q) | 57731 |
| 153275 | Style Transfer (SP) | 394987 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 133 GPixel/s | Pixel Fill Rate | 359 GPixel/s |
| 253 GTexel/s | Texture Fill Rate | 718 GTexel/s |
| 16.2 TFLOPS | FLOPS (FP32) | 46 TFLOPS |
| GDDR6 | Memory Type | GDDR6 |
| 8 GB | Memory Size | 16 GB |
| 1750 MHz | Memory Clock | 2250 MHz |
| 14000 Mbps | Effective Memory Speed | 18000 Mbps |
| 256-bit | Bus | 256-bit |
| No | ECC | No |
| 448 GB/s | Memory Bandwidth | 576 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 4.0 x16 |
| 200 W | TGP | 260 W |
| Samsung | Manufacturing | TSMC |
| 8 nm | Fabrication Process | 5 nm |
| 392 mm² | Die Size | 529 mm² |
| 17 billion | Transistor Count | 57 billion |
| 43.37 MTr/mm² | Transistor Density | 107.75 MTr/mm² |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 2.2 |
| 8.6 | CUDA | — |
| Yes | Ray Tracing | Yes |
| DLSS 2 | DLSS | No |
| 1.4a | DisplayPort | 2.1 |
| Nvidia | Vendor | Amd |
| Discrete | Build | Discrete |
| December 2, 2020 | Released | — |
| Desktop | Case | Desktop |
| Gaming | Purpose | Gaming |
| Mid-range | Segment | High-end |
| Ampere | Architecture | RDNA 3.0 |
| GA104 | GPU Codename | Navi 31 |
| -Intel Core i5 12600Kor above | Recommended CPU | — |
| 11978 | Steel Nomad Lite Score | 21359 |
| 11707 | Time Spy | 22537 |
| 54374 | Solar Bay | 90757 |
| 6971 | Port Royal | 12461 |
| 29568 | Fire Strike | 56212 |
| 24415 | Wild Life Extreme | 41740 |
| 109854 | Night Raid | 169038 |
| 2 IPC | Instructions Per Cycle | 4 IPC |
| 1410 MHz | Base Clock | 1287 MHz |
| 1665 MHz | Boost Clock | 2245 MHz |
| 4864 | Shading Units | 5120 |
| 152 | Texture Mapping Units (TMUs) | 320 |
| 80 | Render Output Units (ROPs) | 160 |
| 38 | Compute Units (Pipelines) | — |
| 152 | Tensor Cores | — |
| 38 | Ray-tracing Cores | 80 |
| 128KB per cluster | L1 Cache | 256KB per cluster |
| 4MB shared | L2 Cache | 6MB shared |