Quick/Comparisons
The Nvidia RTX 5000 Ada takes the lead in gaming performance, thanks to its higher overall score and strong ray tracing capabilities.
The Nvidia RTX 5000 Ada outperforms the Nvidia RTX 4500 Ada in content creation, with a significant 14-point gap in workstation performance.
Both GPUs have similar AI/ML performance, with the Nvidia RTX 4500 Ada and Nvidia RTX 5000 Ada tied at 52 and 54 respectively.
62 | Gaming | 73 |
68 | Workstation | 82 |
52 | AI/ML | 54 |
45 | Energy Efficiency | 45 |
62 | Quick Comparison Final Score | 73 |
| AD103 | Graphics Processor | AD102 |
| 7680 | Cores | 12800 |
| 240 / 80 | TMUs / ROPs | 400 / 176 |
| 5579.97 | Blender GPU | 7348.98 |
| 126 | Forza Horizon 5 | 146 |
| 175 | The Witcher 3 | 186 |
| 213 | Counter-Strike 2 | 213 |
| 145 | Far Cry 6 | 151 |
| 106 | Hogwarts Legacy | 117 |
| 160 | Call of Duty | 179 |
| 99 | Ghost of Tsushima | 108 |
| 134 | Cyberpunk 2077 | 148 |
| 216 | Tomb Raider | 242 |
| 153 | Average FPS | 166 |
| 181 | 1080p High | 196 |
| 153 | 1080p Ultra | 166 |
| 119 | 1440p Ultra | 132 |
| 68 | 4K Ultra | 78 |
| High | Margin of Error | High |
| 206673 | GB6 Compute Score | 260099 |
| 309.9 img/sec | Background Blur | 286.6 img/sec |
| 198.4 img/sec | Face Detection | 220.4 img/sec |
| 7.42 Gpixels/sec | Horizon Detection | 9.93 Gpixels/sec |
| 9.94 Gpixels/sec | Edge Detection | 13.7 Gpixels/sec |
| 12.4 Gpixels/sec | Gaussian Blur | 18.9 Gpixels/sec |
| 2.02 Gpixels/sec | Feature Matching | 2.2 Gpixels/sec |
| 1060 Gpixels/sec | Stereo Matching | 1580 Gpixels/sec |
| 31078.1 FPS | Particle Physics | 41235.6 FPS |
| OpenCL | API | OpenCL |
| 28210 | G3D Mark Score | 30196 |
| 1183 | G2D Mark | 998 |
| 258 FPS | DirectX 11 | 273 FPS |
| 92 FPS | DirectX 12 | 106 FPS |
| 16666 Ops/s | GPU Compute | 20736 Ops/s |
| 26396 | GB6 ML Single Precision | 24715 |
| 40553 | GB6 ML Half Precision | 36753 |
| 20133 | GB6 ML Quantized | 19649 |
| 11193 | Image Classification (SP) | 9943 |
| 20124 | Image Segmentation (HP) | 19696 |
| 34904 | Image Super Resolution (Q) | 32252 |
| 42820 | Face Detection (HP) | 43098 |
| 149991 | Pose Estimation (Q) | 138689 |
| 3789 | Text Classification (SP) | 2893 |
| 4268 | Machine Translation (HP) | 4064 |
| 13068 | Object Detection (SP) | 12467 |
| 47003 | Depth Estimation (Q) | 50352 |
| 300643 | Style Transfer (SP) | 265482 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 206 GPixel/s | Pixel Fill Rate | 449 GPixel/s |
| 619 GTexel/s | Texture Fill Rate | 1020 GTexel/s |
| 39.6 TFLOPS | FLOPS (FP32) | 65.3 TFLOPS |
| GDDR6 | Memory Type | GDDR6 |
| 24 GB | Memory Size | 32 GB |
| 2250 MHz | Memory Clock | 2250 MHz |
| 18000 Mbps | Effective Memory Speed | 18000 Mbps |
| 192-bit | Bus | 256-bit |
| No | ECC | No |
| 432 GB/s | Memory Bandwidth | 576 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 4.0 x16 |
| 210 W | TGP | 250 W |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 5 nm |
| 379 mm² | Die Size | 609 mm² |
| 45 billion | Transistor Count | 76 billion |
| 118.73 MTr/mm² | Transistor Density | 124.79 MTr/mm² |
| 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 |
| 12 | DirectX | 12 |
| 1.3 | Vulkan | 1.3 |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| August 23, 2023 | Released | August 9, 2023 |
| Desktop | Case | Desktop |
| Professional | Purpose | Professional |
| High-end | Segment | High-end |
| Ada Lovelace | Architecture | Ada Lovelace |
| AD103 | GPU Codename | AD102 |
| -Intel Core i7 14700Kor above | Recommended CPU | -Intel Core i9 14900Kor above |
| 12727 | Port Royal | 17754 |
| 21566 | Steel Nomad Lite Score | 29190 |
| 19844 | Time Spy | 18675 |
| 95907 | Solar Bay | 128428 |
| 48020 | Fire Strike | 62027 |
| 40308 | Wild Life Extreme | 52866 |
| — | Night Raid | 210443 |
| 2070 MHz | Base Clock | 1155 MHz |
| 2580 MHz | Boost Clock | 2550 MHz |
| 7680 | Shading Units | 12800 |
| 240 | Texture Mapping Units (TMUs) | 400 |
| 80 | Render Output Units (ROPs) | 176 |
| 60 | Compute Units (Pipelines) | 100 |
| 240 | Tensor Cores | 400 |
| 60 | Ray-tracing Cores | 100 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 48MB shared | L2 Cache | 72MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare Nvidia RTX 4500 with another →