Quick/Comparisons
62 | Gaming | 29 |
68 | Workstation | 39 |
52 | AI/ML |
Compare with something else?
Compare Nvidia RTX 4500 with another →27 |
45 | Energy Efficiency | 66 |
62 | Quick Comparison Final Score | 38 |
| AD103 | Graphics Processor | GB207-300-A1 |
| 7680 | Cores | 2560 |
| 240 / 80 | TMUs / ROPs | 80 / 32 |
| 5579.97 | Blender GPU | 2723.94 |
| 126 | Forza Horizon 5 | 65 |
| 175 | The Witcher 3 | 94 |
| 213 | Counter-Strike 2 | 125 |
| 145 | Far Cry 6 | 79 |
| 106 | Hogwarts Legacy | 53 |
| 160 | Call of Duty | 68 |
| 99 | Ghost of Tsushima | 45 |
| 134 | Cyberpunk 2077 | 60 |
| 216 | Tomb Raider | 94 |
| 153 | Average FPS | 76 |
| 181 | 1080p High | 95 |
| 153 | 1080p Ultra | 76 |
| 119 | 1440p Ultra | 55 |
| 68 | 4K Ultra | 29 |
| High | Margin of Error | High |
| 206673 | GB6 Compute Score | 88051 |
| 309.9 img/sec | Background Blur | 137.7 img/sec |
| 198.4 img/sec | Face Detection | 95 img/sec |
| 7.42 Gpixels/sec | Horizon Detection | 2.99 Gpixels/sec |
| 9.94 Gpixels/sec | Edge Detection | 3.89 Gpixels/sec |
| 12.4 Gpixels/sec | Gaussian Blur | 4.7 Gpixels/sec |
| 2.02 Gpixels/sec | Feature Matching | 1.11 Gpixels/sec |
| 1060 Gpixels/sec | Stereo Matching | 396.8 Gpixels/sec |
| 31078.1 FPS | Particle Physics | 12852.4 FPS |
| OpenCL | API | OpenCL |
| 28210 | G3D Mark Score | 14784 |
| 1183 | G2D Mark | 731 |
| 258 FPS | DirectX 11 | 117 FPS |
| 92 FPS | DirectX 12 | 73 FPS |
| 16666 Ops/s | GPU Compute | 5832 Ops/s |
| 26396 | GB6 ML Single Precision | 14484 |
| 40553 | GB6 ML Half Precision | 29177 |
| 20133 | GB6 ML Quantized | 11633 |
| 11193 | Image Classification (SP) | 6114 |
| 20124 | Image Segmentation (HP) | 17332 |
| 34904 | Image Super Resolution (Q) | 15867 |
| 42820 | Face Detection (HP) | 32890 |
| 149991 | Pose Estimation (Q) | 60048 |
| 3789 | Text Classification (SP) | 2929 |
| 4268 | Machine Translation (HP) | 4631 |
| 13068 | Object Detection (SP) | 7844 |
| 47003 | Depth Estimation (Q) | 26902 |
| 300643 | Style Transfer (SP) | 94600 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 206 GPixel/s | Pixel Fill Rate | 52-77 GPixel/s |
| 619 GTexel/s | Texture Fill Rate | 130-193 GTexel/s |
| 39.6 TFLOPS | FLOPS (FP32) | 8.3-12.3 TFLOPS |
| GDDR6 | Memory Type | GDDR7 |
| 24 GB | Memory Size | 8 GB |
| 2250 MHz | Memory Clock | 1500 MHz |
| 18000 Mbps | Effective Memory Speed | 24000 Mbps |
| 192-bit | Bus | 128-bit |
| No | ECC | No |
| 432 GB/s | Memory Bandwidth | 384 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 5.0 x16 |
| 210 W | TGP | 35-115 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 5 nm |
| 379 mm² | Die Size | 149 mm² |
| 45 billion | Transistor Count | 16.9 billion |
| 118.73 MTr/mm² | Transistor Density | 113.42 MTr/mm² |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 8.9 | CUDA | 12 |
| Yes | Ray Tracing | Yes |
| DLSS 3 | DLSS | DLSS 4 |
| 1.4a | DisplayPort | 2.1b |
| 12 | DirectX | 12.2 |
| 1.3 | Vulkan | 1.4 |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| August 23, 2023 | Released | March 18, 2025 |
| Desktop | Case | Laptop |
| Professional | Purpose | Professional |
| High-end | Segment | Entry-level |
| Ada Lovelace | Architecture | Blackwell 2.0 |
| AD103 | GPU Codename | GB207-300-A1 |
| -Intel Core i7 14700Kor above | Recommended CPU | -Intel Core Ultra 5 245HXor above |
| 12727 | Port Royal | 5015 |
| 21566 | Steel Nomad Lite Score | 8158 |
| 19844 | Time Spy | 8545 |
| 95907 | Solar Bay | 17288 |
| 48020 | Fire Strike | 19480 |
| 40308 | Wild Life Extreme | — |
| — | Night Raid | 89940 |
| 2070 MHz | Base Clock | 1020-1627 MHz (configurable) |
| 2580 MHz | Boost Clock | 1627-2407 MHz (configurable) |
| 7680 | Shading Units | 2560 |
| 240 | Texture Mapping Units (TMUs) | 80 |
| 80 | Render Output Units (ROPs) | 32 |
| 60 | Compute Units (Pipelines) | 20 |
| 240 | Tensor Cores | 80 |
| 60 | Ray-tracing Cores | 20 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 48MB shared | L2 Cache | 32MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |