Quick/Comparisons
The Nvidia RTX 5000 Ada takes the lead in gaming performance, thanks to its higher score of 73. Its strong performance is a result of its Ada Lovelace architecture and 32GB GDDR6 memory.
The Nvidia RTX 5000 Ada is the clear winner in content creation, boasting a score of 82 to the GeForce RTX 4070's 55. Its strong workstation performance is driven by its large 32GB memory pool.
The Nvidia RTX 5000 Ada and GeForce RTX 4070 are nearly tied in AI and machine learning, with the Nvidia RTX 5000 Ada scoring 54 to the GeForce RTX 4070's 48. However, the Nvidia RTX 5000 Ada's score is only 6 points higher, making this a close call.
54 | Gaming | 73 |
55 | Workstation | 82 |
48 | AI/ML | 54 |
44 | Energy Efficiency | 45 |
53 | Quick Comparison Final Score | 73 |
| AD104 | Graphics Processor | AD102 |
| 5888 | Cores | 12800 |
| 184 / 64 | TMUs / ROPs | 400 / 176 |
| 5394.14 | Blender GPU | 7348.98 |
| 114 | Forza Horizon 5 | 146 |
| 163 | The Witcher 3 | 186 |
| 202 | Counter-Strike 2 | 213 |
| 135 | Far Cry 6 | 151 |
| 96 | Hogwarts Legacy | 117 |
| 143 | Call of Duty | 179 |
| 89 | Ghost of Tsushima | 108 |
| 115 | Cyberpunk 2077 | 148 |
| 196 | Tomb Raider | 242 |
| 139 | Average FPS | 166 |
| 165 | 1080p High | 196 |
| 139 | 1080p Ultra | 166 |
| 107 | 1440p Ultra | 132 |
| 63 | 4K Ultra | 78 |
| Low | Margin of Error | High |
| 167522 | GB6 Compute Score | 260099 |
| 233 img/sec | Background Blur | 286.6 img/sec |
| 141.7 img/sec | Face Detection | 220.4 img/sec |
| 6.62 Gpixels/sec | Horizon Detection | 9.93 Gpixels/sec |
| 9.19 Gpixels/sec | Edge Detection | 13.7 Gpixels/sec |
| 9.2 Gpixels/sec | Gaussian Blur | 18.9 Gpixels/sec |
| 1.78 Gpixels/sec | Feature Matching | 2.2 Gpixels/sec |
| 845.4 Gpixels/sec | Stereo Matching | 1580 Gpixels/sec |
| 25043.9 FPS | Particle Physics | 41235.6 FPS |
| OpenCL | API | OpenCL |
| 26921 | G3D Mark Score | 30196 |
| 1168 | G2D Mark | 998 |
| 243 FPS | DirectX 11 | 273 FPS |
| 103 FPS | DirectX 12 | 106 FPS |
| 14623 Ops/s | GPU Compute | 20736 Ops/s |
| 27166 | GB6 ML Single Precision | 24715 |
| 40963 | GB6 ML Half Precision | 36753 |
| 20332 | GB6 ML Quantized | 19649 |
| 10246 | Image Classification (SP) | 9943 |
| 26153 | Image Segmentation (HP) | 19696 |
| 32054 | Image Super Resolution (Q) | 32252 |
| 51337 | Face Detection (HP) | 43098 |
| 124617 | Pose Estimation (Q) | 138689 |
| 3703 | Text Classification (SP) | 2893 |
| 5484 | Machine Translation (HP) | 4064 |
| 13134 | Object Detection (SP) | 12467 |
| 50548 | Depth Estimation (Q) | 50352 |
| 263477 | Style Transfer (SP) | 265482 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 158 GPixel/s | Pixel Fill Rate | 449 GPixel/s |
| 455 GTexel/s | Texture Fill Rate | 1020 GTexel/s |
| 29.1 TFLOPS | FLOPS (FP32) | 65.3 TFLOPS |
| GDDR6X | Memory Type | GDDR6 |
| 12 GB | Memory Size | 32 GB |
| 1313 MHz | Memory Clock | 2250 MHz |
| 21000 Mbps | Effective Memory Speed | 18000 Mbps |
| 192-bit | Bus | 256-bit |
| No | ECC | No |
| 504 GB/s | Memory Bandwidth | 576 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 4.0 x16 |
| 200 W | TGP | 250 W |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 5 nm |
| 294 mm² | Die Size | 609 mm² |
| 35 billion | Transistor Count | 76 billion |
| 119.05 MTr/mm² | Transistor Density | 124.79 MTr/mm² |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 8.9 | CUDA | 8.9 |
| Yes | Ray Tracing | Yes |
| DLSS 3 | DLSS | DLSS 3 |
| 1.4a | DisplayPort | 1.4a |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| April 13, 2023 | Released | August 9, 2023 |
| Desktop | Case | Desktop |
| Gaming | Purpose | Professional |
| Mid-range | Segment | High-end |
| Ada Lovelace | Architecture | Ada Lovelace |
| AD104 | GPU Codename | AD102 |
| -Intel Core i7 14700Kor above | Recommended CPU | -Intel Core i9 14900Kor above |
| 18089 | Steel Nomad Lite Score | 29190 |
| 17806 | Time Spy | 18675 |
| 82784 | Solar Bay | 128428 |
| 11154 | Port Royal | 17754 |
| 44075 | Fire Strike | 62027 |
| 35470 | Wild Life Extreme | 52866 |
| 153508 | Night Raid | 210443 |
| 1920 MHz | Base Clock | 1155 MHz |
| 2475 MHz | Boost Clock | 2550 MHz |
| 5888 | Shading Units | 12800 |
| 184 | Texture Mapping Units (TMUs) | 400 |
| 64 | Render Output Units (ROPs) | 176 |
| 46 | Compute Units (Pipelines) | 100 |
| 184 | Tensor Cores | 400 |
| 46 | Ray-tracing Cores | 100 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 36MB shared | L2 Cache | 72MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare GeForce RTX 4070 with another →