Quick/Comparisons
22 | Gaming | 16 |
28 | Workstation | 10 |
16 | AI/ML |
Compare with something else?
Compare GeForce GTX 1070 with another →7 |
20 | Energy Efficiency | 94 |
23 | Quick Comparison Final Score | 26 |
| GP104 | Graphics Processor | — |
| 2432 | Cores | 512 |
| 152 / 64 | TMUs / ROPs | 32 / 16 |
| 629.59 | Blender GPU | — |
| 55 | Forza Horizon 5 | — |
| 77 | The Witcher 3 | — |
| 108 | Counter-Strike 2 | — |
| 70 | Far Cry 6 | — |
| 42 | Hogwarts Legacy | — |
| 53 | Call of Duty | — |
| 39 | Ghost of Tsushima | — |
| 45 | Cyberpunk 2077 | — |
| 80 | Tomb Raider | — |
| 63 | Average FPS | — |
| 79 | 1080p High | — |
| 63 | 1080p Ultra | — |
| 45 | 1440p Ultra | — |
| 24 | 4K Ultra | — |
| Low | Margin of Error | — |
| 57735 | GB6 Compute Score | 16321 |
| 149.2 img/sec | Background Blur | 37.6 img/sec |
| 67.6 img/sec | Face Detection | 17.2 img/sec |
| 2.36 Gpixels/sec | Horizon Detection | 0.3 Gpixels/sec |
| 3.57 Gpixels/sec | Edge Detection | 0.63 Gpixels/sec |
| 4.32 Gpixels/sec | Gaussian Blur | 0.94 Gpixels/sec |
| 0.27 Gpixels/sec | Feature Matching | 0.22 Gpixels/sec |
| 209.6 Gpixels/sec | Stereo Matching | 60 Gpixels/sec |
| 6584.9 FPS | Particle Physics | 3695.9 FPS |
| OpenCL | API | OpenCL |
| 14658 | G3D Mark Score | — |
| 876 | G2D Mark | — |
| 107 FPS | DirectX 11 | — |
| 52 FPS | DirectX 12 | — |
| 7162 Ops/s | GPU Compute | — |
| 8949 | GB6 ML Single Precision | 3770 |
| 8769 | GB6 ML Half Precision | 5146 |
| 6546 | GB6 ML Quantized | 2824 |
| 3473 | Image Classification (SP) | 2902 |
| 5149 | Image Segmentation (HP) | 5473 |
| 9722 | Image Super Resolution (Q) | 2967 |
| 9306 | Face Detection (HP) | 6068 |
| 34870 | Pose Estimation (Q) | 8220 |
| 2139 | Text Classification (SP) | 609 |
| 2424 | Machine Translation (HP) | 1146 |
| 4566 | Object Detection (SP) | 2766 |
| 14316 | Depth Estimation (Q) | 6405 |
| 88984 | Style Transfer (SP) | 26796 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 108 GPixel/s | Pixel Fill Rate | 28-32 GPixel/s |
| 256 GTexel/s | Texture Fill Rate | 56-64 GTexel/s |
| 8.2 TFLOPS | FLOPS (FP32) | 1.8-2 TFLOPS |
| GDDR5 | Memory Type | System Shared |
| 8 GB | Memory Size | — |
| 2002 MHz | Memory Clock | — |
| 8000 Mbps | Effective Memory Speed | — |
| 256-bit | Bus | — |
| No | ECC | — |
| 256.3 GB/s | Memory Bandwidth | — |
| PCIe 3.0 x16 | Interface | Ring Bus |
| 180 W | TGP | 14 W |
| TSMC | Manufacturing | TSMC |
| 16 nm | Fabrication Process | 5 nm |
| 314 mm² | Die Size | — |
| 7 billion | Transistor Count | — |
| 22.29 MTr/mm² | Transistor Density | — |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 6.1 | CUDA | — |
| No | Ray Tracing | Yes |
| No | DLSS | No |
| 1.4a | DisplayPort | 2.1 |
| Nvidia | Vendor | Intel |
| Discrete | Build | Integrated |
| November 2, 2017 | Released | December 14, 2023 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Generic |
| Mid-range | Segment | Mid-range |
| Pascal | Architecture | Xe LPG |
| GP104 | GPU Codename | — |
| -Intel Core i7 7700Kor above | Recommended CPU | -Intel Core Ultra 5 125Uor above |
| 6671 | Steel Nomad Lite Score | — |
| 6843 | Time Spy | — |
| 18267 | Solar Bay | — |
| 1552 | Port Royal | — |
| 20115 | Fire Strike | — |
| 13248 | Wild Life Extreme | — |
| 78890 | Night Raid | — |
| 1607 MHz | Base Clock | 300 MHz |
| 1683 MHz | Boost Clock | 1750-2000 MHz (configurable) |
| 2432 | Shading Units | 512 |
| 152 | Texture Mapping Units (TMUs) | 32 |
| 64 | Render Output Units (ROPs) | 16 |
| 19 | Compute Units (Pipelines) | 64 |
| No | Tensor Cores | — |
| No | Ray-tracing Cores | 4 |
| 48KB per cluster | L1 Cache | 768KB shared |
| 2MB shared | L2 Cache | — |
| 2 IPC | Instructions Per Cycle | 2 IPC |