Quick/Comparisons
43 | Gaming | 5 |
44 | Workstation | 11 |
35 | AI/ML |
Compare with something else?
Compare GeForce RTX 3070 with another →4 |
35 | Energy Efficiency | 89 |
42 | Quick Comparison Final Score | 20 |
| GA104 | Graphics Processor | Mendocino |
| 5888 | Cores | 128 |
| 184 / 96 | TMUs / ROPs | 8 / 4 |
| 3368.42 | Blender GPU | 47.24 |
| 99 | Forza Horizon 5 | — |
| 143 | The Witcher 3 | — |
| 194 | Counter-Strike 2 | — |
| 121 | Far Cry 6 | — |
| 80 | Hogwarts Legacy | — |
| 113 | Call of Duty | — |
| 74 | Ghost of Tsushima | — |
| 94 | Cyberpunk 2077 | — |
| 153 | Tomb Raider | — |
| 119 | Average FPS | — |
| 141 | 1080p High | — |
| 119 | 1080p Ultra | — |
| 88 | 1440p Ultra | — |
| 50 | 4K Ultra | — |
| Low | Margin of Error | — |
| 126195 | GB6 Compute Score | 5099 |
| 161.6 img/sec | Background Blur | 12 img/sec |
| 117.2 img/sec | Face Detection | 6.42 img/sec |
| 5.89 Gpixels/sec | Horizon Detection | 0.16 Gpixels/sec |
| 8.05 Gpixels/sec | Edge Detection | 0.23 Gpixels/sec |
| 6.65 Gpixels/sec | Gaussian Blur | 0.16 Gpixels/sec |
| 1.25 Gpixels/sec | Feature Matching | 0.07 Gpixels/sec |
| 550.6 Gpixels/sec | Stereo Matching | 18.2 Gpixels/sec |
| 17613.1 FPS | Particle Physics | 828.6 FPS |
| OpenCL | API | OpenCL |
| 22128 | G3D Mark Score | 1810 |
| 1005 | G2D Mark | 536 |
| 182 FPS | DirectX 11 | 17 FPS |
| 84 FPS | DirectX 12 | 7 FPS |
| 11062 Ops/s | GPU Compute | 771 Ops/s |
| 19789 | GB6 ML Single Precision | 1917 |
| 33013 | GB6 ML Half Precision | 2053 |
| 12443 | GB6 ML Quantized | 1571 |
| 7624 | Image Classification (SP) | 660 |
| 23420 | Image Segmentation (HP) | 627 |
| 17898 | Image Super Resolution (Q) | 2053 |
| 40535 | Face Detection (HP) | 2153 |
| 83624 | Pose Estimation (Q) | 5787 |
| 2724 | Text Classification (SP) | 713 |
| 4481 | Machine Translation (HP) | 1114 |
| 10304 | Object Detection (SP) | 709 |
| 30418 | Depth Estimation (Q) | 3233 |
| 180089 | Style Transfer (SP) | 11091 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 166 GPixel/s | Pixel Fill Rate | 8-9 GPixel/s |
| 317 GTexel/s | Texture Fill Rate | 15-18 GTexel/s |
| 20.3 TFLOPS | FLOPS (FP32) | 0.5-0.6 TFLOPS |
| GDDR6 | Memory Type | System Shared |
| 8 GB | Memory Size | — |
| 1750 MHz | Memory Clock | — |
| 14000 Mbps | Effective Memory Speed | — |
| 256-bit | Bus | — |
| No | ECC | No |
| 448 GB/s | Memory Bandwidth | — |
| PCIe 4.0 x16 | Interface | PCIe 4.0 x8 |
| 220 W | TGP | 15 W |
| Samsung | Manufacturing | TSMC |
| 8 nm | Fabrication Process | 6 nm |
| 392 mm² | Die Size | 122 mm² |
| 17 billion | Transistor Count | 3.4 billion |
| 43.37 MTr/mm² | Transistor Density | 27.87 MTr/mm² |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 2 |
| 8.6 | CUDA | — |
| Yes | Ray Tracing | Yes |
| DLSS 2 | DLSS | No |
| 1.4a | DisplayPort | — |
| Nvidia | Vendor | Amd |
| Discrete | Build | Integrated |
| October 22, 2020 | Released | October 1, 2022 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Generic |
| Mid-range | Segment | Entry-level |
| Ampere | Architecture | RDNA 2.0 |
| GA104 | GPU Codename | Mendocino |
| -Intel Core i7 12700Kor above | Recommended CPU | -AMD Ryzen 5 7520Uor above |
| 14153 | Steel Nomad Lite Score | 447 |
| 13643 | Time Spy | 519 |
| 64426 | Solar Bay | 2078 |
| 8311 | Port Royal | 309 |
| 34054 | Fire Strike | 1760 |
| 28486 | Wild Life Extreme | 994 |
| 122039 | Night Raid | 7447 |
| 1500 MHz | Base Clock | 400 MHz |
| 1725 MHz | Boost Clock | 1900-2200 MHz (configurable) |
| 5888 | Shading Units | 128 |
| 184 | Texture Mapping Units (TMUs) | 8 |
| 96 | Render Output Units (ROPs) | 4 |
| 46 | Compute Units (Pipelines) | 2 |
| 184 | Tensor Cores | — |
| 46 | Ray-tracing Cores | 2 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 4MB shared | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |