Quick/Comparisons
The GeForce RTX 4090's superior performance is due to its 92 score, significantly outpacing the GeForce GTX 1070 Ti's 22 score.
The GeForce RTX 4090's 91 score surpasses the GeForce GTX 1070 Ti's 28 score, making it the better choice for content creation.
The GeForce RTX 4090's 76 score outperforms the GeForce GTX 1070 Ti's 16 score, indicating its superiority in AI/ML compute.
22 | Gaming | 92 |
28 | Workstation | 91 |
16 | AI/ML | 76 |
20 | Energy Efficiency | 43 |
23 | Quick Comparison Final Score | 89 |
| GP104 | Graphics Processor | AD102 |
| 2432 | Cores | 16384 |
| 152 / 64 | TMUs / ROPs | 512 / 176 |
| 629.59 | Blender GPU | 11688.48 |
| 55 | Forza Horizon 5 | 203 |
| 77 | The Witcher 3 | 270 |
| 108 | Counter-Strike 2 | 245 |
| 70 | Far Cry 6 | 183 |
| 42 | Hogwarts Legacy | 165 |
| 53 | Call of Duty | 269 |
| 39 | Ghost of Tsushima | 150 |
| 45 | Cyberpunk 2077 | 204 |
| 80 | Tomb Raider | 283 |
| 63 | Average FPS | 219 |
| 79 | 1080p High | 256 |
| 63 | 1080p Ultra | 219 |
| 45 | 1440p Ultra | 176 |
| 24 | 4K Ultra | 107 |
| Low | Margin of Error | Low |
| 57735 | GB6 Compute Score | 316301 |
| 149.2 img/sec | Background Blur | 318.9 img/sec |
| 67.6 img/sec | Face Detection | 222.8 img/sec |
| 2.36 Gpixels/sec | Horizon Detection | 14.2 Gpixels/sec |
| 3.57 Gpixels/sec | Edge Detection | 21 Gpixels/sec |
| 4.32 Gpixels/sec | Gaussian Blur | 23.9 Gpixels/sec |
| 0.27 Gpixels/sec | Feature Matching | 2.38 Gpixels/sec |
| 209.6 Gpixels/sec | Stereo Matching | 1950 Gpixels/sec |
| 6584.9 FPS | Particle Physics | 48090.4 FPS |
| OpenCL | API | OpenCL |
| 14658 | G3D Mark Score | 38071 |
| 876 | G2D Mark | 1303 |
| 107 FPS | DirectX 11 | 324 FPS |
| 52 FPS | DirectX 12 | 151 FPS |
| 7162 Ops/s | GPU Compute | 26193 Ops/s |
| 8949 | GB6 ML Single Precision | 42962 |
| 8769 | GB6 ML Half Precision | 59636 |
| 6546 | GB6 ML Quantized | 31670 |
| 3473 | Image Classification (SP) | 15600 |
| 5149 | Image Segmentation (HP) | 36657 |
| 9722 | Image Super Resolution (Q) | 48114 |
| 9306 | Face Detection (HP) | 78111 |
| 34870 | Pose Estimation (Q) | 289639 |
| 2139 | Text Classification (SP) | 4074 |
| 2424 | Machine Translation (HP) | 6349 |
| 4566 | Object Detection (SP) | 21186 |
| 14316 | Depth Estimation (Q) | 75599 |
| 88984 | Style Transfer (SP) | 601428 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 108 GPixel/s | Pixel Fill Rate | 444 GPixel/s |
| 256 GTexel/s | Texture Fill Rate | 1290 GTexel/s |
| 8.2 TFLOPS | FLOPS (FP32) | 82.6 TFLOPS |
| GDDR5 | Memory Type | GDDR6X |
| 8 GB | Memory Size | 24 GB |
| 2002 MHz | Memory Clock | 2625 MHz |
| 8000 Mbps | Effective Memory Speed | 21000 Mbps |
| 256-bit | Bus | 384-bit |
| No | ECC | No |
| 256.3 GB/s | Memory Bandwidth | 1010 GB/s |
| PCIe 3.0 x16 | Interface | PCIe 4.0 x16 |
| 180 W | TGP | 450 W |
| TSMC | Manufacturing | TSMC |
| 16 nm | Fabrication Process | 5 nm |
| 314 mm² | Die Size | 609 mm² |
| 7 billion | Transistor Count | 76.3 billion |
| 22.29 MTr/mm² | Transistor Density | 125.29 MTr/mm² |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 6.1 | CUDA | 8.9 |
| No | Ray Tracing | Yes |
| No | DLSS | DLSS 3 |
| 1.4a | DisplayPort | 1.4a |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| November 2, 2017 | Released | September 20, 2022 |
| Desktop | Case | Desktop |
| Gaming | Purpose | Gaming |
| Mid-range | Segment | High-end |
| Pascal | Architecture | Ada Lovelace |
| GP104 | GPU Codename | AD102 |
| -Intel Core i7 7700Kor above | Recommended CPU | -Intel Core i9 14900Kor above |
| 6671 | Steel Nomad Lite Score | 42169 |
| 6843 | Time Spy | 36311 |
| 18267 | Solar Bay | 187386 |
| 1552 | Port Royal | 26128 |
| 20115 | Fire Strike | 72387 |
| 13248 | Wild Life Extreme | 85230 |
| 78890 | Night Raid | 195880 |
| 1607 MHz | Base Clock | 2235 MHz |
| 1683 MHz | Boost Clock | 2520 MHz |
| 2432 | Shading Units | 16384 |
| 152 | Texture Mapping Units (TMUs) | 512 |
| 64 | Render Output Units (ROPs) | 176 |
| 19 | Compute Units (Pipelines) | 128 |
| No | Tensor Cores | 512 |
| No | Ray-tracing Cores | 128 |
| 48KB per cluster | L1 Cache | 128KB per cluster |
| 2MB shared | L2 Cache | 72MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare GeForce GTX 1070 with another →