Quick/Comparisons
66 | Gaming | 31 |
66 | Workstation | 32 |
57 | AI/ML |
Compare with something else?
Compare GeForce RTX 4070 with another →31 |
44 | Energy Efficiency | 65 |
64 | Quick Comparison Final Score | 37 |
| AD104 | Graphics Processor | Navi 33 |
| 7680 | Cores | 2048 |
| 240 / 80 | TMUs / ROPs | 128 / 64 |
| 6639.7 | Blender GPU | 1021.44 |
| 141 | Forza Horizon 5 | 78 |
| 182 | The Witcher 3 | 103 |
| 216 | Counter-Strike 2 | 137 |
| 148 | Far Cry 6 | 95 |
| 114 | Hogwarts Legacy | 58 |
| 171 | Call of Duty | 89 |
| 106 | Ghost of Tsushima | 63 |
| 142 | Cyberpunk 2077 | 65 |
| 245 | Tomb Raider | 114 |
| 163 | Average FPS | 89 |
| 194 | 1080p High | 112 |
| 163 | 1080p Ultra | 89 |
| 128 | 1440p Ultra | 62 |
| 75 | 4K Ultra | 34 |
| Low | Margin of Error | Low |
| 219163 | GB6 Compute Score | 75137 |
| 371.2 img/sec | Background Blur | 183.6 img/sec |
| 235 img/sec | Face Detection | 91.6 img/sec |
| 7.51 Gpixels/sec | Horizon Detection | 2.68 Gpixels/sec |
| 10.6 Gpixels/sec | Edge Detection | 3.95 Gpixels/sec |
| 12.4 Gpixels/sec | Gaussian Blur | 3.39 Gpixels/sec |
| 2.03 Gpixels/sec | Feature Matching | 0.84 Gpixels/sec |
| 1120 Gpixels/sec | Stereo Matching | 244.3 Gpixels/sec |
| 30420.2 FPS | Particle Physics | 9216.4 FPS |
| OpenCL | API | OpenCL |
| 31568 | G3D Mark Score | 15485 |
| 1209 | G2D Mark | 765 |
| 288 FPS | DirectX 11 | 137 FPS |
| 116 FPS | DirectX 12 | 67 FPS |
| 18221 Ops/s | GPU Compute | 7016 Ops/s |
| 32405 | GB6 ML Single Precision | 18185 |
| 48989 | GB6 ML Half Precision | 23280 |
| 24894 | GB6 ML Quantized | 14842 |
| 12496 | Image Classification (SP) | 8076 |
| 33378 | Image Segmentation (HP) | 10203 |
| 37180 | Image Super Resolution (Q) | 22390 |
| 62919 | Face Detection (HP) | 29334 |
| 171125 | Pose Estimation (Q) | 84068 |
| 3630 | Text Classification (SP) | 3066 |
| 6420 | Machine Translation (HP) | 4439 |
| 15804 | Object Detection (SP) | 7062 |
| 58647 | Depth Estimation (Q) | 38916 |
| 331379 | Style Transfer (SP) | 160714 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 209 GPixel/s | Pixel Fill Rate | 118-141 GPixel/s |
| 626 GTexel/s | Texture Fill Rate | 236-282 GTexel/s |
| 40.1 TFLOPS | FLOPS (FP32) | 15.1-18 TFLOPS |
| GDDR6X | Memory Type | GDDR6 |
| 12 GB | Memory Size | 8 GB |
| 1313 MHz | Memory Clock | 2250 MHz |
| 21000 Mbps | Effective Memory Speed | 18000 Mbps |
| 192-bit | Bus | 128-bit |
| No | ECC | — |
| 504 GB/s | Memory Bandwidth | 288 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 4.0 x8 |
| 285 W | TGP | 75-100 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 6 nm |
| 294 mm² | Die Size | 204 mm² |
| 35 billion | Transistor Count | 13.3 billion |
| 119.05 MTr/mm² | Transistor Density | 65.2 MTr/mm² |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 2.2 |
| 8.9 | CUDA | — |
| Yes | Ray Tracing | Yes |
| DLSS 3 | DLSS | No |
| 1.4a | DisplayPort | 2.1 |
| Mid-range | Segment | High-end |
| Ada Lovelace | Architecture | RDNA 3.0 |
| AD104 | GPU Codename | Navi 33 |
| -Intel Core i7 14700Kor above | Recommended CPU | -AMD Ryzen 7 7735HSor above |
| Nvidia | Vendor | Amd |
| Discrete | Build | Discrete |
| January 3, 2023 | Released | January 4, 2023 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Gaming |
| 23618 | Steel Nomad Lite Score | 9302 |
| 22751 | Time Spy | 10279 |
| 108422 | Solar Bay | 41160 |
| 13994 | Port Royal | 5350 |
| 53484 | Fire Strike | 28706 |
| 45397 | Wild Life Extreme | 17962 |
| 171258 | Night Raid | 101381 |
| 2310 MHz | Base Clock | 1500 MHz |
| 2610 MHz | Boost Clock | 1840-2200 MHz (configurable) |
| 7680 | Shading Units | 2048 |
| 240 | Texture Mapping Units (TMUs) | 128 |
| 80 | Render Output Units (ROPs) | 64 |
| 60 | Compute Units (Pipelines) | 32 |
| 240 | Tensor Cores | — |
| 60 | Ray-tracing Cores | 32 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 48MB shared | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 4 IPC |