Quick/Comparisons
The GeForce RTX 5070 takes the lead in gaming performance, thanks to its higher score of 64. Its 12GB GDDR7 memory also provides a significant advantage over the GeForce RTX 2070's 8GB GDDR6 memory.
The GeForce RTX 5070 outperforms the GeForce RTX 2070 in content creation, with a score of 61 compared to the GeForce RTX 2070's 37. This is likely due to the GeForce RTX 5070's more efficient architecture and increased VRAM.
The GeForce RTX 5070 excels in AI and machine learning tasks, with a score of 54 compared to the GeForce RTX 2070's 28. This is likely due to the GeForce RTX 5070's dedicated AI tensor cores and more advanced architecture.
29 | Gaming | 64 |
37 | Workstation | 61 |
28 | AI/ML |
Compare with something else?
Compare GeForce RTX 2070 with another →54 |
28 | Energy Efficiency | 49 |
31 | Quick Comparison Final Score | 61 |
| TU106 | Graphics Processor | GB205-300 |
| 2304 | Cores | 6144 |
| 144 / 64 | TMUs / ROPs | 192 / 80 |
| 2200.46 | Blender GPU | 6160.42 |
| 66 | Forza Horizon 5 | 137 |
| 97 | The Witcher 3 | 187 |
| 134 | Counter-Strike 2 | 218 |
| 85 | Far Cry 6 | 146 |
| 56 | Hogwarts Legacy | 116 |
| 70 | Call of Duty | 176 |
| 50 | Ghost of Tsushima | 108 |
| 62 | Cyberpunk 2077 | 140 |
| 99 | Tomb Raider | 236 |
| 80 | Average FPS | 163 |
| 99 | 1080p High | 193 |
| 80 | 1080p Ultra | 163 |
| 58 | 1440p Ultra | 127 |
| 31 | 4K Ultra | 75 |
| Low | Margin of Error | Low |
| 97923 | GB6 Compute Score | 190967 |
| 200 img/sec | Background Blur | 254.4 img/sec |
| 106.8 img/sec | Face Detection | 180.6 img/sec |
| 3.73 Gpixels/sec | Horizon Detection | 7.91 Gpixels/sec |
| 4.57 Gpixels/sec | Edge Detection | 10.2 Gpixels/sec |
| 4.84 Gpixels/sec | Gaussian Blur | 10.7 Gpixels/sec |
| 1.07 Gpixels/sec | Feature Matching | 1.84 Gpixels/sec |
| 452 Gpixels/sec | Stereo Matching | 975.7 Gpixels/sec |
| 11141.9 FPS | Particle Physics | 27816.5 FPS |
| OpenCL | API | OpenCL |
| 16066 | G3D Mark Score | 28767 |
| 817 | G2D Mark | 1300 |
| 130 FPS | DirectX 11 | 280 FPS |
| 61 FPS | DirectX 12 | 106 FPS |
| 6388 Ops/s | GPU Compute | 14507 Ops/s |
| 12633 | GB6 ML Single Precision | 30221 |
| 20355 | GB6 ML Half Precision | 53037 |
| 8999 | GB6 ML Quantized | 23474 |
| 5244 | Image Classification (SP) | 12270 |
| 9644 | Image Segmentation (HP) | 30237 |
| 13716 | Image Super Resolution (Q) | 35585 |
| 21326 | Face Detection (HP) | 62662 |
| 57284 | Pose Estimation (Q) | 145190 |
| 1857 | Text Classification (SP) | 4198 |
| 2542 | Machine Translation (HP) | 6233 |
| 6466 | Object Detection (SP) | 16709 |
| 22978 | Depth Estimation (Q) | 58300 |
| 111896 | Style Transfer (SP) | 246598 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 104 GPixel/s | Pixel Fill Rate | 201 GPixel/s |
| 233 GTexel/s | Texture Fill Rate | 482 GTexel/s |
| 7.5 TFLOPS | FLOPS (FP32) | 30.9 TFLOPS |
| GDDR6 | Memory Type | GDDR7 |
| 8 GB | Memory Size | 12 GB |
| 1750 MHz | Memory Clock | 1750 MHz |
| 14000 Mbps | Effective Memory Speed | 28000 Mbps |
| 256-bit | Bus | 192-bit |
| No | ECC | No |
| 448 GB/s | Memory Bandwidth | 672 GB/s |
| PCIe 3.0 x16 | Interface | PCIe 5.0 x16 |
| 175 W | TGP | 250 W |
| TSMC | Manufacturing | TSMC |
| 12 nm | Fabrication Process | 4 nm |
| 445 mm² | Die Size | 263 mm² |
| 10 billion | Transistor Count | 31.1 billion |
| 22.47 MTr/mm² | Transistor Density | 118.25 MTr/mm² |
| 12 | DirectX | 12.2 |
| 1.3 | Vulkan | 1.4 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 7.5 | CUDA | 12 |
| Yes | Ray Tracing | Yes |
| DLSS 2 | DLSS | DLSS 4 |
| 1.4a | DisplayPort | 2.1b |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| October 17, 2018 | Released | January 7, 2025 |
| Desktop | Case | Desktop |
| Gaming | Purpose | Gaming |
| Mid-range | Segment | Mid-range |
| Turing | Architecture | Blackwell 2.0 |
| TU106 | GPU Codename | GB205-300 |
| -Intel Core i7 9700Kor above | Recommended CPU | -Intel Core Ultra 5 245Kor above |
| 8488 | Steel Nomad Lite Score | 23229 |
| 9076 | Time Spy | 22678 |
| 39691 | Solar Bay | 106731 |
| 5194 | Port Royal | 14423 |
| 23209 | Fire Strike | 58244 |
| 18260 | Wild Life Extreme | 43724 |
| 92325 | Night Raid | 174956 |
| 1410 MHz | Base Clock | 2160 MHz |
| 1620 MHz | Boost Clock | 2512 MHz |
| 2304 | Shading Units | 6144 |
| 144 | Texture Mapping Units (TMUs) | 192 |
| 64 | Render Output Units (ROPs) | 80 |
| 36 | Compute Units (Pipelines) | 48 |
| 288 | Tensor Cores | 192 |
| 36 | Ray-tracing Cores | 48 |
| 64KB per cluster | L1 Cache | 128KB per cluster |
| 4MB shared | L2 Cache | 48MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |