Quick/Comparisons
61 | Gaming | 37 |
62 | Workstation | 59 |
54 | AI/ML |
Compare with something else?
Compare GeForce RTX 4070 with another →45 |
46 | Energy Efficiency | 59 |
60 | Quick Comparison Final Score | 47 |
| AD104 | Graphics Processor | GB205-300-A1 |
| 7168 | Cores | 5888 |
| 224 / 80 | TMUs / ROPs | 184 / 80 |
| 6153.24 | Blender GPU | 3405.25 |
| 131 | Forza Horizon 5 | 79 |
| 176 | The Witcher 3 | 111 |
| 212 | Counter-Strike 2 | 149 |
| 145 | Far Cry 6 | 94 |
| 106 | Hogwarts Legacy | 63 |
| 165 | Call of Duty | 83 |
| 100 | Ghost of Tsushima | 58 |
| 131 | Cyberpunk 2077 | 73 |
| 226 | Tomb Raider | 112 |
| 155 | Average FPS | 91 |
| 183 | 1080p High | 112 |
| 155 | 1080p Ultra | 91 |
| 120 | 1440p Ultra | 66 |
| 70 | 4K Ultra | 37 |
| Low | Margin of Error | High |
| 206598 | GB6 Compute Score | 167869 |
| 359.5 img/sec | Background Blur | 238.8 img/sec |
| 210 img/sec | Face Detection | 147.6 img/sec |
| 7.61 Gpixels/sec | Horizon Detection | 6.73 Gpixels/sec |
| 10.4 Gpixels/sec | Edge Detection | 9.27 Gpixels/sec |
| 11.3 Gpixels/sec | Gaussian Blur | 10.7 Gpixels/sec |
| 1.97 Gpixels/sec | Feature Matching | 1.68 Gpixels/sec |
| 969.3 Gpixels/sec | Stereo Matching | 778.4 Gpixels/sec |
| 29061.4 FPS | Particle Physics | 23023.5 FPS |
| OpenCL | API | OpenCL |
| 29964 | G3D Mark Score | 21959 |
| 1193 | G2D Mark | 833 |
| 271 FPS | DirectX 11 | 224 FPS |
| 110 FPS | DirectX 12 | 100 FPS |
| 16985 Ops/s | GPU Compute | 7578 Ops/s |
| 30706 | GB6 ML Single Precision | 24473 |
| 43851 | GB6 ML Half Precision | 44496 |
| 23776 | GB6 ML Quantized | 18924 |
| 12146 | Image Classification (SP) | 9790 |
| 23932 | Image Segmentation (HP) | 28672 |
| 36321 | Image Super Resolution (Q) | 26553 |
| 52836 | Face Detection (HP) | 53161 |
| 179788 | Pose Estimation (Q) | 120306 |
| 3371 | Text Classification (SP) | 3406 |
| 5017 | Machine Translation (HP) | 5910 |
| 15164 | Object Detection (SP) | 12573 |
| 62500 | Depth Estimation (Q) | 47957 |
| 357489 | Style Transfer (SP) | 199379 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 198 GPixel/s | Pixel Fill Rate | 128-194 GPixel/s |
| 554 GTexel/s | Texture Fill Rate | 295-447 GTexel/s |
| 35.5 TFLOPS | FLOPS (FP32) | 18.9-28.6 TFLOPS |
| GDDR6X | Memory Type | GDDR7 |
| 12 GB | Memory Size | 12 GB |
| 1313 MHz | Memory Clock | 1750 MHz |
| 21000 Mbps | Effective Memory Speed | 28000 Mbps |
| 192-bit | Bus | 192-bit |
| No | ECC | No |
| 504.2 GB/s | Memory Bandwidth | 672 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 5.0 x16 |
| 220 W | TGP | 60-140 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 5 nm |
| 294 mm² | Die Size | 263 mm² |
| 35 billion | Transistor Count | 31.1 billion |
| 119.05 MTr/mm² | Transistor Density | 118.25 MTr/mm² |
| 12 | DirectX | 12.2 |
| 1.3 | Vulkan | 1.4 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 8.9 | CUDA | 12 |
| Yes | Ray Tracing | Yes |
| DLSS 3 | DLSS | DLSS 4 |
| 1.4a | DisplayPort | 2.1b |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| January 17, 2024 | Released | March 18, 2025 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Professional |
| Mid-range | Segment | Mid-range |
| Ada Lovelace | Architecture | Blackwell 2.0 |
| AD104 | GPU Codename | GB205-300-A1 |
| -Intel Core i7 14700Kor above | Recommended CPU | -Intel Core Ultra 7 265HXor above |
| 21568 | Steel Nomad Lite Score | 8100 |
| 20882 | Time Spy | 12327 |
| 98166 | Solar Bay | 69295 |
| 13122 | Port Royal | — |
| 50092 | Fire Strike | 28339 |
| 41630 | Wild Life Extreme | — |
| 164777 | Night Raid | — |
| 1980 MHz | Base Clock | 848-1605 MHz (configurable) |
| 2475 MHz | Boost Clock | 1605-2430 MHz (configurable) |
| 7168 | Shading Units | 5888 |
| 224 | Texture Mapping Units (TMUs) | 184 |
| 80 | Render Output Units (ROPs) | 80 |
| 56 | Compute Units (Pipelines) | 46 |
| 224 | Tensor Cores | 184 |
| 56 | Ray-tracing Cores | 46 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 48MB shared | L2 Cache | 48MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |