Quick/Comparisons
54 | Gaming | 32 |
55 | Workstation | 33 |
48 | AI/ML |
Compare with something else?
Compare GeForce RTX 4070 with another →41 |
44 | Energy Efficiency | 66 |
53 | Quick Comparison Final Score | 39 |
| AD104 | Graphics Processor | Strix Halo |
| 5888 | Cores | 2560 |
| 184 / 64 | TMUs / ROPs | 160 / 64 |
| 5394.14 | Blender GPU | 1335.18 |
| 114 | Forza Horizon 5 | 77 |
| 163 | The Witcher 3 | 101 |
| 202 | Counter-Strike 2 | 138 |
| 135 | Far Cry 6 | 92 |
| 96 | Hogwarts Legacy | 61 |
| 143 | Call of Duty | 89 |
| 89 | Ghost of Tsushima | 64 |
| 115 | Cyberpunk 2077 | 65 |
| 196 | Tomb Raider | 115 |
| 139 | Average FPS | 89 |
| 165 | 1080p High | 114 |
| 139 | 1080p Ultra | 89 |
| 107 | 1440p Ultra | 64 |
| 63 | 4K Ultra | 34 |
| Low | Margin of Error | Low |
| 167522 | GB6 Compute Score | 93515 |
| 233 img/sec | Background Blur | 232.1 img/sec |
| 141.7 img/sec | Face Detection | 122.5 img/sec |
| 6.62 Gpixels/sec | Horizon Detection | 2.92 Gpixels/sec |
| 9.19 Gpixels/sec | Edge Detection | 4.07 Gpixels/sec |
| 9.2 Gpixels/sec | Gaussian Blur | 3.49 Gpixels/sec |
| 1.78 Gpixels/sec | Feature Matching | 1.08 Gpixels/sec |
| 845.4 Gpixels/sec | Stereo Matching | 396.3 Gpixels/sec |
| 25043.9 FPS | Particle Physics | 13016.5 FPS |
| OpenCL | API | OpenCL |
| 26921 | G3D Mark Score | 18115 |
| 1168 | G2D Mark | 1172 |
| 243 FPS | DirectX 11 | 157 FPS |
| 103 FPS | DirectX 12 | 84 FPS |
| 14623 Ops/s | GPU Compute | 9062 Ops/s |
| 27166 | GB6 ML Single Precision | 22565 |
| 40963 | GB6 ML Half Precision | 30098 |
| 20332 | GB6 ML Quantized | 18100 |
| 10246 | Image Classification (SP) | 11176 |
| 26153 | Image Segmentation (HP) | 14027 |
| 32054 | Image Super Resolution (Q) | 27198 |
| 51337 | Face Detection (HP) | 40613 |
| 124617 | Pose Estimation (Q) | 92878 |
| 3703 | Text Classification (SP) | 4158 |
| 5484 | Machine Translation (HP) | 6473 |
| 13134 | Object Detection (SP) | 7431 |
| 50548 | Depth Estimation (Q) | 41825 |
| 263477 | Style Transfer (SP) | 190221 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 158 GPixel/s | Pixel Fill Rate | 186 GPixel/s |
| 455 GTexel/s | Texture Fill Rate | 464 GTexel/s |
| 29.1 TFLOPS | FLOPS (FP32) | 14.8 TFLOPS |
| GDDR6X | Memory Type | System Shared |
| 12 GB | Memory Size | — |
| 1313 MHz | Memory Clock | — |
| 21000 Mbps | Effective Memory Speed | — |
| 192-bit | Bus | — |
| No | ECC | — |
| 504 GB/s | Memory Bandwidth | — |
| PCIe 4.0 x16 | Interface | PCIe 5.0 x16 |
| 200 W | TGP | 45-120 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 4 nm |
| 294 mm² | Die Size | 308 mm² |
| 35 billion | Transistor Count | — |
| 119.05 MTr/mm² | Transistor Density | — |
| 12 | DirectX | 12.2 |
| 1.3 | Vulkan | 1.4 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 2.1 |
| 8.9 | CUDA | No |
| Yes | Ray Tracing | Yes |
| DLSS 3 | DLSS | No |
| 1.4a | DisplayPort | 2.1a |
| Nvidia | Vendor | Amd |
| Discrete | Build | Integrated |
| April 13, 2023 | Released | January 7, 2025 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Generic |
| Mid-range | Segment | Mid-range |
| Ada Lovelace | Architecture | RDNA 3.5 |
| AD104 | GPU Codename | Strix Halo |
| -Intel Core i7 14700Kor above | Recommended CPU | -AMD Ryzen AI Max+ 395or above |
| 18089 | Steel Nomad Lite Score | 9818 |
| 17806 | Time Spy | 10010 |
| 82784 | Solar Bay | 46934 |
| 11154 | Port Royal | 5512 |
| 44075 | Fire Strike | 26722 |
| 35470 | Wild Life Extreme | 19043 |
| 153508 | Night Raid | 100428 |
| 1920 MHz | Base Clock | 1295 MHz |
| 2475 MHz | Boost Clock | 2900 MHz |
| 5888 | Shading Units | 2560 |
| 184 | Texture Mapping Units (TMUs) | 160 |
| 64 | Render Output Units (ROPs) | 64 |
| 46 | Compute Units (Pipelines) | 40 |
| 184 | Tensor Cores | No |
| 46 | Ray-tracing Cores | 40 |
| 128KB per cluster | L1 Cache | 512KB per cluster |
| 36MB shared | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |