Quick/Comparisons
44 | Gaming | 28 |
48 | Workstation | 29 |
39 | AI/ML |
Compare with something else?
Compare GeForce RTX 2080 with another →33 |
27 | Energy Efficiency | 66 |
44 | Quick Comparison Final Score | 35 |
| TU102 | Graphics Processor | Strix Halo |
| 4352 | Cores | 2048 |
| 272 / 88 | TMUs / ROPs | 128 / 64 |
| 3046.64 | Blender GPU | 1188.98 |
| 101 | Forza Horizon 5 | 70 |
| 144 | The Witcher 3 | 94 |
| 199 | Counter-Strike 2 | 123 |
| 122 | Far Cry 6 | 85 |
| 82 | Hogwarts Legacy | 55 |
| 118 | Call of Duty | 83 |
| 78 | Ghost of Tsushima | 56 |
| 98 | Cyberpunk 2077 | 61 |
| 155 | Tomb Raider | 103 |
| 122 | Average FPS | 81 |
| 145 | 1080p High | 104 |
| 122 | 1080p Ultra | 81 |
| 92 | 1440p Ultra | 57 |
| 52 | 4K Ultra | 31 |
| Low | Margin of Error | Medium |
| 144745 | GB6 Compute Score | 81081 |
| 211 img/sec | Background Blur | 201.1 img/sec |
| 141.3 img/sec | Face Detection | 104.8 img/sec |
| 6.24 Gpixels/sec | Horizon Detection | 2.53 Gpixels/sec |
| 8.1 Gpixels/sec | Edge Detection | 3.37 Gpixels/sec |
| 9.22 Gpixels/sec | Gaussian Blur | 3.12 Gpixels/sec |
| 1.46 Gpixels/sec | Feature Matching | 0.93 Gpixels/sec |
| 627.3 Gpixels/sec | Stereo Matching | 327.5 Gpixels/sec |
| 16980.5 FPS | Particle Physics | 12215.6 FPS |
| OpenCL | API | OpenCL |
| 21477 | G3D Mark Score | 16166 |
| 927 | G2D Mark | 1109 |
| 192 FPS | DirectX 11 | 148 FPS |
| 83 FPS | DirectX 12 | 70 FPS |
| 9906 Ops/s | GPU Compute | 7513 Ops/s |
| 19878 | GB6 ML Single Precision | 18145 |
| 31237 | GB6 ML Half Precision | 22099 |
| 12663 | GB6 ML Quantized | 14884 |
| 8213 | Image Classification (SP) | 9259 |
| 19891 | Image Segmentation (HP) | 11219 |
| 17598 | Image Super Resolution (Q) | 19775 |
| 39785 | Face Detection (HP) | 31987 |
| 84003 | Pose Estimation (Q) | 62399 |
| 2062 | Text Classification (SP) | 3757 |
| 4192 | Machine Translation (HP) | 6121 |
| 10444 | Object Detection (SP) | 6979 |
| 31856 | Depth Estimation (Q) | 35735 |
| 207773 | Style Transfer (SP) | 129996 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 136 GPixel/s | Pixel Fill Rate | 179 GPixel/s |
| 420 GTexel/s | Texture Fill Rate | 358 GTexel/s |
| 13.4 TFLOPS | FLOPS (FP32) | 11.5 TFLOPS |
| GDDR6 | Memory Type | System Shared |
| 11 GB | Memory Size | — |
| 1750 MHz | Memory Clock | — |
| 14000 Mbps | Effective Memory Speed | — |
| 352-bit | Bus | — |
| No | ECC | — |
| 616 GB/s | Memory Bandwidth | — |
| PCIe 3.0 x16 | Interface | PCIe 5.0 x16 |
| 250 W | TGP | 45-120 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 12 nm | Fabrication Process | 4 nm |
| 754 mm² | Die Size | 308 mm² |
| 18 billion | Transistor Count | — |
| 23.87 MTr/mm² | Transistor Density | — |
| 12 | DirectX | 12.2 |
| 1.3 | Vulkan | 1.4 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 2.1 |
| 7.5 | CUDA | No |
| Yes | Ray Tracing | Yes |
| DLSS 2 | DLSS | No |
| 1.4a | DisplayPort | 2.1a |
| Nvidia | Vendor | Amd |
| Discrete | Build | Integrated |
| September 27, 2018 | Released | January 7, 2025 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Generic |
| High-end | Segment | Mid-range |
| Turing | Architecture | RDNA 3.5 |
| TU102 | GPU Codename | Strix Halo |
| -Intel Core i9 9900Kor above | Recommended CPU | -AMD Ryzen AI Max 385or above |
| 14284 | Steel Nomad Lite Score | 8570 |
| 14656 | Time Spy | 9171 |
| 64076 | Solar Bay | 41112 |
| 9109 | Port Royal | 5181 |
| 35666 | Fire Strike | 24607 |
| 29554 | Wild Life Extreme | 18225 |
| 115500 | Night Raid | 98347 |
| 1350 MHz | Base Clock | 1295 MHz |
| 1545 MHz | Boost Clock | 2800 MHz |
| 4352 | Shading Units | 2048 |
| 272 | Texture Mapping Units (TMUs) | 128 |
| 88 | Render Output Units (ROPs) | 64 |
| 68 | Compute Units (Pipelines) | 32 |
| 544 | Tensor Cores | No |
| 68 | Ray-tracing Cores | 32 |
| 64KB per cluster | L1 Cache | 512KB per cluster |
| 5632KB shared | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |