Quick/Comparisons
The GeForce RTX 5080's superior gaming performance is due to its Blackwell 2.0 architecture and 16GB GDDR7 memory, outscoring the GeForce GTX 1660 Ti by 67 points.
The GeForce RTX 5080 takes the lead in content creation with a 56-point advantage, thanks to its strong workstation performance and PCIe 5.0 x16 interface.
The GeForce RTX 5080's capable AI/ML compute and 16GB GDDR7 memory give it a 52-point lead over the GeForce GTX 1660 Ti in AI and machine learning tasks.
20 | Gaming | 87 |
29 | Workstation | 85 |
18 | AI/ML | 70 |
44 | Energy Efficiency | 49 |
23 | Quick Comparison Final Score | 84 |
| TU116 | Graphics Processor | GB203-400 |
| 1536 | Cores | 10752 |
| 96 / 48 | TMUs / ROPs | 336 / 112 |
| 796.71 | Blender GPU | 9138.54 |
| 52 | Forza Horizon 5 | 185 |
| 69 | The Witcher 3 | 237 |
| 96 | Counter-Strike 2 | 237 |
| 64 | Far Cry 6 | 176 |
| 37 | Hogwarts Legacy | 154 |
| 47 | Call of Duty | 241 |
| 32 | Ghost of Tsushima | 137 |
| 42 | Cyberpunk 2077 | 190 |
| 74 | Tomb Raider | 270 |
| 57 | Average FPS | 203 |
| 74 | 1080p High | 239 |
| 57 | 1080p Ultra | 203 |
| 41 | 1440p Ultra | 164 |
| 21 | 4K Ultra | 97 |
| Low | Margin of Error | Low |
| 72340 | GB6 Compute Score | 297747 |
| 124.8 img/sec | Background Blur | 319.6 img/sec |
| 69.4 img/sec | Face Detection | 225 img/sec |
| 2.67 Gpixels/sec | Horizon Detection | 13.6 Gpixels/sec |
| 3.48 Gpixels/sec | Edge Detection | 19.5 Gpixels/sec |
| 3.52 Gpixels/sec | Gaussian Blur | 20.4 Gpixels/sec |
| 0.92 Gpixels/sec | Feature Matching | 2.33 Gpixels/sec |
| 355.9 Gpixels/sec | Stereo Matching | 1750 Gpixels/sec |
| 9094.4 FPS | Particle Physics | 43676.6 FPS |
| OpenCL | API | OpenCL |
| 12661 | G3D Mark Score | 35695 |
| 779 | G2D Mark | 1407 |
| 103 FPS | DirectX 11 | 327 FPS |
| 51 FPS | DirectX 12 | 144 FPS |
| 5074 Ops/s | GPU Compute | 19397 Ops/s |
| 7620 | GB6 ML Quantized | 29206 |
| 4272 | Image Classification (SP) | 15441 |
| 6435 | Image Segmentation (HP) | 47173 |
| 9635 | Image Super Resolution (Q) | 43223 |
| 10867 | Face Detection (HP) | 92863 |
| 41061 | Pose Estimation (Q) | 202202 |
| 1737 | Text Classification (SP) | 3701 |
| 3301 | Machine Translation (HP) | 7905 |
| 5600 | Object Detection (SP) | 20955 |
| 18914 | Depth Estimation (Q) | 69068 |
| 83427 | Style Transfer (SP) | 349205 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 10788 | GB6 ML Single Precision | 38994 |
| 8753 | GB6 ML Half Precision | 66609 |
| 85 GPixel/s | Pixel Fill Rate | 293 GPixel/s |
| 170 GTexel/s | Texture Fill Rate | 879 GTexel/s |
| 5.4 TFLOPS | FLOPS (FP32) | 56.3 TFLOPS |
| GDDR6 | Memory Type | GDDR7 |
| 6 GB | Memory Size | 16 GB |
| 1500 MHz | Memory Clock | 1875 MHz |
| 12000 Mbps | Effective Memory Speed | 30000 Mbps |
| 192-bit | Bus | 256-bit |
| No | ECC | No |
| 288 GB/s | Memory Bandwidth | 960 GB/s |
| PCIe 3.0 x16 | Interface | PCIe 5.0 x16 |
| 120 W | TGP | 360 W |
| TSMC | Manufacturing | TSMC |
| 12 nm | Fabrication Process | 4 nm |
| 284 mm² | Die Size | 378 mm² |
| 6 billion | Transistor Count | 45.6 billion |
| 21.13 MTr/mm² | Transistor Density | 120.63 MTr/mm² |
| 12 | DirectX | 12.2 |
| 1.3 | Vulkan | 1.4 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 7.5 | CUDA | 12 |
| No | Ray Tracing | Yes |
| No | DLSS | DLSS 4 |
| 1.4a | DisplayPort | 2.1b |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| February 22, 2019 | Released | January 7, 2025 |
| Desktop | Case | Desktop |
| Gaming | Purpose | Gaming |
| Entry-level | Segment | High-end |
| Turing | Architecture | Blackwell 2.0 |
| TU116 | GPU Codename | GB203-400 |
| -Intel Core i5 9500or above | Recommended CPU | -Intel Core Ultra 7 265Kor above |
| 5944 | Steel Nomad Lite Score | 36200 |
| 6339 | Time Spy | 32989 |
| 20534 | Solar Bay | 160901 |
| 1650 | Port Royal | 23257 |
| 16653 | Fire Strike | 75518 |
| 13005 | Wild Life Extreme | 69544 |
| 76361 | Night Raid | 200077 |
| 1500 MHz | Base Clock | 2295 MHz |
| 1770 MHz | Boost Clock | 2617 MHz |
| 1536 | Shading Units | 10752 |
| 96 | Texture Mapping Units (TMUs) | 336 |
| 48 | Render Output Units (ROPs) | 112 |
| 24 | Compute Units (Pipelines) | 84 |
| No | Tensor Cores | 336 |
| No | Ray-tracing Cores | 84 |
| 64KB per cluster | L1 Cache | 128KB per cluster |
| 1536KB shared | L2 Cache | 64MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare GeForce GTX 1660 with another →