Quick/Comparisons
22 | Gaming | 44 |
28 | Workstation | 48 |
16 | AI/ML |
Compare with something else?
Compare GeForce GTX 1070 with another →39 |
20 | Energy Efficiency | 27 |
23 | Quick Comparison Final Score | 44 |
| GP104 | Graphics Processor | TU102 |
| 2432 | Cores | 4352 |
| 152 / 64 | TMUs / ROPs | 272 / 88 |
| 629.59 | Blender GPU | 3046.64 |
| 55 | Forza Horizon 5 | 101 |
| 77 | The Witcher 3 | 144 |
| 108 | Counter-Strike 2 | 199 |
| 70 | Far Cry 6 | 122 |
| 42 | Hogwarts Legacy | 82 |
| 53 | Call of Duty | 118 |
| 39 | Ghost of Tsushima | 78 |
| 45 | Cyberpunk 2077 | 98 |
| 80 | Tomb Raider | 155 |
| 63 | Average FPS | 122 |
| 79 | 1080p High | 145 |
| 63 | 1080p Ultra | 122 |
| 45 | 1440p Ultra | 92 |
| 24 | 4K Ultra | 52 |
| Low | Margin of Error | Low |
| 57735 | GB6 Compute Score | 144745 |
| 149.2 img/sec | Background Blur | 211 img/sec |
| 67.6 img/sec | Face Detection | 141.3 img/sec |
| 2.36 Gpixels/sec | Horizon Detection | 6.24 Gpixels/sec |
| 3.57 Gpixels/sec | Edge Detection | 8.1 Gpixels/sec |
| 4.32 Gpixels/sec | Gaussian Blur | 9.22 Gpixels/sec |
| 0.27 Gpixels/sec | Feature Matching | 1.46 Gpixels/sec |
| 209.6 Gpixels/sec | Stereo Matching | 627.3 Gpixels/sec |
| 6584.9 FPS | Particle Physics | 16980.5 FPS |
| OpenCL | API | OpenCL |
| 14658 | G3D Mark Score | 21477 |
| 876 | G2D Mark | 927 |
| 107 FPS | DirectX 11 | 192 FPS |
| 52 FPS | DirectX 12 | 83 FPS |
| 7162 Ops/s | GPU Compute | 9906 Ops/s |
| 8949 | GB6 ML Single Precision | 19878 |
| 8769 | GB6 ML Half Precision | 31237 |
| 6546 | GB6 ML Quantized | 12663 |
| 3473 | Image Classification (SP) | 8213 |
| 5149 | Image Segmentation (HP) | 19891 |
| 9722 | Image Super Resolution (Q) | 17598 |
| 9306 | Face Detection (HP) | 39785 |
| 34870 | Pose Estimation (Q) | 84003 |
| 2139 | Text Classification (SP) | 2062 |
| 2424 | Machine Translation (HP) | 4192 |
| 4566 | Object Detection (SP) | 10444 |
| 14316 | Depth Estimation (Q) | 31856 |
| 88984 | Style Transfer (SP) | 207773 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 108 GPixel/s | Pixel Fill Rate | 136 GPixel/s |
| 256 GTexel/s | Texture Fill Rate | 420 GTexel/s |
| 8.2 TFLOPS | FLOPS (FP32) | 13.4 TFLOPS |
| GDDR5 | Memory Type | GDDR6 |
| 8 GB | Memory Size | 11 GB |
| 2002 MHz | Memory Clock | 1750 MHz |
| 8000 Mbps | Effective Memory Speed | 14000 Mbps |
| 256-bit | Bus | 352-bit |
| No | ECC | No |
| 256.3 GB/s | Memory Bandwidth | 616 GB/s |
| PCIe 3.0 x16 | Interface | PCIe 3.0 x16 |
| 180 W | TGP | 250 W |
| TSMC | Manufacturing | TSMC |
| 16 nm | Fabrication Process | 12 nm |
| 314 mm² | Die Size | 754 mm² |
| 7 billion | Transistor Count | 18 billion |
| 22.29 MTr/mm² | Transistor Density | 23.87 MTr/mm² |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 6.1 | CUDA | 7.5 |
| No | Ray Tracing | Yes |
| No | DLSS | DLSS 2 |
| 1.4a | DisplayPort | 1.4a |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| November 2, 2017 | Released | September 27, 2018 |
| Desktop | Case | Desktop |
| Gaming | Purpose | Gaming |
| Mid-range | Segment | High-end |
| Pascal | Architecture | Turing |
| GP104 | GPU Codename | TU102 |
| -Intel Core i7 7700Kor above | Recommended CPU | -Intel Core i9 9900Kor above |
| 6671 | Steel Nomad Lite Score | 14284 |
| 6843 | Time Spy | 14656 |
| 18267 | Solar Bay | 64076 |
| 1552 | Port Royal | 9109 |
| 20115 | Fire Strike | 35666 |
| 13248 | Wild Life Extreme | 29554 |
| 78890 | Night Raid | 115500 |
| 1607 MHz | Base Clock | 1350 MHz |
| 1683 MHz | Boost Clock | 1545 MHz |
| 2432 | Shading Units | 4352 |
| 152 | Texture Mapping Units (TMUs) | 272 |
| 64 | Render Output Units (ROPs) | 88 |
| 19 | Compute Units (Pipelines) | 68 |
| No | Tensor Cores | 544 |
| No | Ray-tracing Cores | 68 |
| 48KB per cluster | L1 Cache | 64KB per cluster |
| 2MB shared | L2 Cache | 5632KB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |