Quick/Comparisons
The Radeon RX 7900 XTX's superior gaming performance is due to its higher benchmark scores and larger 24GB memory pool, making it the clear winner for gaming use cases.
The GeForce RTX 3060 Ti's slightly higher workstation score and dedicated AI tensor cores make it the winner for content creation use cases, despite the Radeon RX 7900 XTX's strong showing.
The Radeon RX 7900 XTX's strong AI/ML compute capabilities and higher score make it a strong contender, but the score gap is only 43 points, so we call this use case a tie.
38 | Gaming | 81 |
42 | Workstation | 75 |
31 | AI/ML | 74 |
35 | Energy Efficiency | 43 |
38 | Quick Comparison Final Score | 77 |
| GA104 | Graphics Processor | Navi 31 |
| 4864 | Cores | 6144 |
| 152 / 80 | TMUs / ROPs | 384 / 192 |
| 2992.04 | Blender GPU | 4055.35 |
| 87 | Forza Horizon 5 | 174 |
| 124 | The Witcher 3 | 256 |
| 167 | Counter-Strike 2 | 232 |
| 106 | Far Cry 6 | 182 |
| 71 | Hogwarts Legacy | 137 |
| 100 | Call of Duty | 255 |
| 68 | Ghost of Tsushima | 138 |
| 82 | Cyberpunk 2077 | 181 |
| 128 | Tomb Raider | 277 |
| 104 | Average FPS | 204 |
| 127 | 1080p High | 241 |
| 104 | 1080p Ultra | 204 |
| 77 | 1440p Ultra | 159 |
| 43 | 4K Ultra | 94 |
| Low | Margin of Error | Low |
| 115518 | GB6 Compute Score | 224686 |
| 185.1 img/sec | Background Blur | 423.9 img/sec |
| 108.4 img/sec | Face Detection | 270.8 img/sec |
| 5.35 Gpixels/sec | Horizon Detection | 9.23 Gpixels/sec |
| 7.17 Gpixels/sec | Edge Detection | 13.7 Gpixels/sec |
| 5.54 Gpixels/sec | Gaussian Blur | 11.1 Gpixels/sec |
| 1.15 Gpixels/sec | Feature Matching | 1.74 Gpixels/sec |
| 478 Gpixels/sec | Stereo Matching | 871 Gpixels/sec |
| 15183 FPS | Particle Physics | 29891.2 FPS |
| OpenCL | API | OpenCL |
| 20272 | G3D Mark Score | 31406 |
| 990 | G2D Mark | 1285 |
| 164 FPS | DirectX 11 | 356 FPS |
| 78 FPS | DirectX 12 | 133 FPS |
| 9905 Ops/s | GPU Compute | 17768 Ops/s |
| 17811 | GB6 ML Single Precision | 42092 |
| 29969 | GB6 ML Half Precision | 50673 |
| 11367 | GB6 ML Quantized | 32874 |
| 7384 | Image Classification (SP) | 18608 |
| 19320 | Image Segmentation (HP) | 25385 |
| 16108 | Image Super Resolution (Q) | 45641 |
| 34615 | Face Detection (HP) | 71321 |
| 74811 | Pose Estimation (Q) | 257258 |
| 2638 | Text Classification (SP) | 5240 |
| 4136 | Machine Translation (HP) | 7425 |
| 8718 | Object Detection (SP) | 21335 |
| 25946 | Depth Estimation (Q) | 66140 |
| 153275 | Style Transfer (SP) | 483867 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 133 GPixel/s | Pixel Fill Rate | 480 GPixel/s |
| 253 GTexel/s | Texture Fill Rate | 959 GTexel/s |
| 16.2 TFLOPS | FLOPS (FP32) | 61.4 TFLOPS |
| GDDR6 | Memory Type | GDDR6 |
| 8 GB | Memory Size | 24 GB |
| 1750 MHz | Memory Clock | 2500 MHz |
| 14000 Mbps | Effective Memory Speed | 20000 Mbps |
| 256-bit | Bus | 384-bit |
| No | ECC | No |
| 448 GB/s | Memory Bandwidth | 960 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 4.0 x16 |
| 200 W | TGP | 355 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 | November 3, 2022 |
| 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 | -Intel Core i7 13700Kor above |
| 11978 | Steel Nomad Lite Score | 29870 |
| 11707 | Time Spy | 30766 |
| 54374 | Solar Bay | 123758 |
| 6971 | Port Royal | 16749 |
| 29568 | Fire Strike | 73217 |
| 24415 | Wild Life Extreme | 58530 |
| 109854 | Night Raid | 184767 |
| 2 IPC | Instructions Per Cycle | 4 IPC |
| 1410 MHz | Base Clock | 1929 MHz |
| 1665 MHz | Boost Clock | 2498 MHz |
| 4864 | Shading Units | 6144 |
| 152 | Texture Mapping Units (TMUs) | 384 |
| 80 | Render Output Units (ROPs) | 192 |
| 38 | Compute Units (Pipelines) | — |
| 152 | Tensor Cores | — |
| 38 | Ray-tracing Cores | 96 |
| 128KB per cluster | L1 Cache | 256KB per cluster |
| 4MB shared | L2 Cache | 6MB shared |
Compare with something else?
Compare GeForce RTX 3060 with another →