Quick/Comparisons
46 | Gaming | 66 |
49 | Workstation | 79 |
39 | AI/ML |
Compare with something else?
Compare GeForce RTX 3070 with another →62 |
35 | Energy Efficiency | 49 |
46 | Quick Comparison Final Score | 66 |
| GA104 | Graphics Processor | GB202-300-A1 |
| 6144 | Cores | 10496 |
| 192 / 96 | TMUs / ROPs | 328 / 112 |
| 3746.83 | Blender GPU | 7401.59 |
| 104 | Forza Horizon 5 | 135 |
| 143 | The Witcher 3 | 180 |
| 198 | Counter-Strike 2 | 214 |
| 125 | Far Cry 6 | 148 |
| 85 | Hogwarts Legacy | 114 |
| 120 | Call of Duty | 175 |
| 81 | Ghost of Tsushima | 106 |
| 102 | Cyberpunk 2077 | 145 |
| 161 | Tomb Raider | 245 |
| 124 | Average FPS | 162 |
| 150 | 1080p High | 191 |
| 124 | 1080p Ultra | 162 |
| 93 | 1440p Ultra | 127 |
| 54 | 4K Ultra | 75 |
| Low | Margin of Error | High |
| 136096 | GB6 Compute Score | 246458 |
| 197 img/sec | Background Blur | 267.1 img/sec |
| 128.1 img/sec | Face Detection | 183.2 img/sec |
| 6.56 Gpixels/sec | Horizon Detection | 11.5 Gpixels/sec |
| 8.81 Gpixels/sec | Edge Detection | 15.8 Gpixels/sec |
| 6.47 Gpixels/sec | Gaussian Blur | 18.1 Gpixels/sec |
| 1.26 Gpixels/sec | Feature Matching | 2.01 Gpixels/sec |
| 609.3 Gpixels/sec | Stereo Matching | 1330 Gpixels/sec |
| 18205 FPS | Particle Physics | 35463.8 FPS |
| OpenCL | API | OpenCL |
| 23260 | G3D Mark Score | 24043 |
| 1061 | G2D Mark | 843 |
| 192 FPS | DirectX 11 | 254 FPS |
| 91 FPS | DirectX 12 | 109 FPS |
| 11452 Ops/s | GPU Compute | 8247 Ops/s |
| 22516 | GB6 ML Single Precision | 31330 |
| 36377 | GB6 ML Half Precision | 53746 |
| 14215 | GB6 ML Quantized | 24267 |
| 8615 | Image Classification (SP) | 12454 |
| 24659 | Image Segmentation (HP) | 33371 |
| 20631 | Image Super Resolution (Q) | 35476 |
| 43834 | Face Detection (HP) | 67998 |
| 98347 | Pose Estimation (Q) | 168369 |
| 3162 | Text Classification (SP) | 3491 |
| 4911 | Machine Translation (HP) | 6115 |
| 11498 | Object Detection (SP) | 16115 |
| 34719 | Depth Estimation (Q) | 59893 |
| 196833 | Style Transfer (SP) | 279891 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 170 GPixel/s | Pixel Fill Rate | 188-202 GPixel/s |
| 340 GTexel/s | Texture Fill Rate | 551-593 GTexel/s |
| 21.7 TFLOPS | FLOPS (FP32) | 35.3-37.9 TFLOPS |
| GDDR6X | Memory Type | GDDR7 |
| 8 GB | Memory Size | 24 GB |
| 1188 MHz | Memory Clock | 1750 MHz |
| 19000 Mbps | Effective Memory Speed | 28000 Mbps |
| 256-bit | Bus | 256-bit |
| No | ECC | — |
| 608.3 GB/s | Memory Bandwidth | 896 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 5.0 x16 |
| 290 W | TGP | 95-175 W (configurable) |
| Samsung | Manufacturing | TSMC |
| 8 nm | Fabrication Process | 5 nm |
| 392 mm² | Die Size | 378 mm² |
| 17 billion | Transistor Count | 45.6 billion |
| 43.37 MTr/mm² | Transistor Density | 120.63 MTr/mm² |
| 12 | DirectX | 12.2 |
| 1.3 | Vulkan | 1.4 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 8.6 | CUDA | 12 |
| Yes | Ray Tracing | Yes |
| DLSS 2 | DLSS | DLSS 4 |
| 1.4a | DisplayPort | 2.1b |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| June 10, 2021 | Released | March 18, 2025 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Professional |
| Mid-range | Segment | High-end |
| Ampere | Architecture | Blackwell 2.0 |
| GA104 | GPU Codename | GB202-300-A1 |
| -Intel Core i7 12700Kor above | Recommended CPU | -Intel Core Ultra 9 275HXor above |
| 15402 | Steel Nomad Lite Score | 24118 |
| 14880 | Time Spy | 21548 |
| 69900 | Solar Bay | 96355 |
| 8884 | Port Royal | 14090 |
| 37238 | Fire Strike | 47636 |
| 31911 | Wild Life Extreme | — |
| 134012 | Night Raid | — |
| 1575 MHz | Base Clock | 990-1680 MHz (configurable) |
| 1770 MHz | Boost Clock | 1680-1807 MHz (configurable) |
| 6144 | Shading Units | 10496 |
| 192 | Texture Mapping Units (TMUs) | 328 |
| 96 | Render Output Units (ROPs) | 112 |
| 48 | Compute Units (Pipelines) | 82 |
| 192 | Tensor Cores | 328 |
| 48 | Ray-tracing Cores | 82 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 4MB shared | L2 Cache | 64MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |