Quick/Comparisons
The GeForce RTX 5090 outperforms the GeForce RTX 3080 Ti Laptop by a significant margin in gaming, thanks to its superior architecture and 32GB GDDR7 memory. Its 100 score is a testament to its exceptional gaming capabilities.
While the GeForce RTX 3080 Ti Laptop holds its own in content creation, the GeForce RTX 5090's 93 score edges it out, thanks to its Blackwell 2.0 architecture and PCIe 5.0 interface.
The GeForce RTX 5090 takes the lead in AI/ML compute, boasting a 95 score, while the GeForce RTX 3080 Ti Laptop trails behind with a 37 score, highlighting the 5090's capabilities in this area.
42 | Gaming | 100 |
48 | Workstation | 93 |
37 | AI/ML | 95 |
37 | Energy Efficiency | 49 |
42 | Quick Comparison Final Score | 96 |
| GA103 (GN20-E8) | Graphics Processor | GB202-300 |
| 7424 | Cores | 21760 |
| 232 / 96 | TMUs / ROPs | 680 / 176 |
| 3502.46 | Blender GPU | 15010.34 |
| 92 | Forza Horizon 5 | 244 |
| 136 | The Witcher 3 | 323 |
| 181 | Counter-Strike 2 | 265 |
| 111 | Far Cry 6 | 202 |
| 79 | Hogwarts Legacy | 207 |
| 106 | Call of Duty | 327 |
| 72 | Ghost of Tsushima | 179 |
| 86 | Cyberpunk 2077 | 229 |
| 143 | Tomb Raider | 299 |
| 112 | Average FPS | 253 |
| 135 | 1080p High | 294 |
| 112 | 1080p Ultra | 253 |
| 83 | 1440p Ultra | 207 |
| 47 | 4K Ultra | 124 |
| Low | Margin of Error | Low |
| 133659 | GB6 Compute Score | 419309 |
| 200.9 img/sec | Background Blur | 333 img/sec |
| 131.2 img/sec | Face Detection | 298.5 img/sec |
| 6.49 Gpixels/sec | Horizon Detection | 21.7 Gpixels/sec |
| 8.9 Gpixels/sec | Edge Detection | 33.7 Gpixels/sec |
| 6.34 Gpixels/sec | Gaussian Blur | 34.5 Gpixels/sec |
| 1.2 Gpixels/sec | Feature Matching | 2.53 Gpixels/sec |
| 571 Gpixels/sec | Stereo Matching | 2850 Gpixels/sec |
| 17240.3 FPS | Particle Physics | 59146.9 FPS |
| OpenCL | API | OpenCL |
| 18948 | G3D Mark Score | 38955 |
| 817 | G2D Mark | 1413 |
| 161 FPS | DirectX 11 | 341 FPS |
| 87 FPS | DirectX 12 | 178 FPS |
| 8302 Ops/s | GPU Compute | 24632 Ops/s |
| 18415 | GB6 ML Single Precision | 53861 |
| 29828 | GB6 ML Half Precision | 81080 |
| 11425 | GB6 ML Quantized | 38938 |
| 6797 | Image Classification (SP) | 18948 |
| 20808 | Image Segmentation (HP) | 55109 |
| 15194 | Image Super Resolution (Q) | 58478 |
| 37084 | Face Detection (HP) | 114685 |
| 82268 | Pose Estimation (Q) | 344729 |
| 2863 | Text Classification (SP) | 4464 |
| 4135 | Machine Translation (HP) | 8184 |
| 8559 | Object Detection (SP) | 27802 |
| 28330 | Depth Estimation (Q) | 85572 |
| 184129 | Style Transfer (SP) | 687080 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 108-153 GPixel/s | Pixel Fill Rate | 424 GPixel/s |
| 261-369 GTexel/s | Texture Fill Rate | 1637 GTexel/s |
| 16.7-23.6 TFLOPS | FLOPS (FP32) | 104.8 TFLOPS |
| GDDR6 | Memory Type | GDDR7 |
| 16 GB | Memory Size | 32 GB |
| 2000 MHz | Memory Clock | 1750 MHz |
| 16000 Mbps | Effective Memory Speed | 28000 Mbps |
| 256-bit | Bus | 512-bit |
| No | ECC | No |
| 512 GB/s | Memory Bandwidth | 1792 GB/s |
| PCIe 4.0 x16 | Interface | PCIe 5.0 x16 |
| 80-175 W (configurable) | TGP | 575 W |
| Samsung | Manufacturing | TSMC |
| 8 nm | Fabrication Process | 4 nm |
| 496 mm² | Die Size | 750 mm² |
| 22 billion | Transistor Count | 92.2 billion |
| 44.35 MTr/mm² | Transistor Density | 122.93 MTr/mm² |
| 12 | DirectX | 12.2 |
| 1.2 | Vulkan | 1.4 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 8.6 | CUDA | 12 |
| Yes | Ray Tracing | Yes |
| DLSS 2 | DLSS | DLSS 4 |
| 1.4a | DisplayPort | 2.1b |
| Nvidia | Vendor | Nvidia |
| Discrete | Build | Discrete |
| January 1, 2022 | Released | January 7, 2025 |
| Laptop | Case | Desktop |
| Gaming | Purpose | Gaming |
| High-end | Segment | High-end |
| Ampere | Architecture | Blackwell 2.0 |
| GA103 (GN20-E8) | GPU Codename | GB202-300 |
| -Intel Core i9 12900Hor above | Recommended CPU | -Intel Core Ultra 9 285Kor above |
| 12913 | Steel Nomad Lite Score | 52638 |
| 12864 | Time Spy | 46941 |
| 56305 | Solar Bay | 235224 |
| 8092 | Port Royal | 38113 |
| 30858 | Fire Strike | 87433 |
| 26307 | Wild Life Extreme | 108736 |
| 98582 | Night Raid | 207058 |
| 585-1230 MHz (configurable) | Base Clock | 2017 MHz |
| 1125-1590 MHz (configurable) | Boost Clock | 2407 MHz |
| 7424 | Shading Units | 21760 |
| 232 | Texture Mapping Units (TMUs) | 680 |
| 96 | Render Output Units (ROPs) | 176 |
| 58 | Compute Units (Pipelines) | 170 |
| 232 | Tensor Cores | 680 |
| 58 | Ray-tracing Cores | 170 |
| 128KB per cluster | L1 Cache | 128KB per cluster |
| 4MB shared | L2 Cache | 96MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare GeForce RTX 3080 with another →