Quick/Comparisons
The GeForce RTX 5090's perfect gaming score is unmatched by the Intel Arc A750, thanks to its superior performance in this area.
The GeForce RTX 5090 takes the lead in content creation due to its strong workstation performance and overall higher score.
The GeForce RTX 5090's high AI/ML score is significantly higher than the Intel Arc A750's, thanks to its capable compute capabilities.
100 | Gaming | 31 |
93 | Workstation | 38 |
95 | AI/ML | 27 |
49 | Energy Efficiency | 43 |
96 | Quick Comparison Final Score | 33 |
| GB202-300 | Graphics Processor | DG2-512 |
| 21760 | Cores | 3584 |
| 680 / 176 | TMUs / ROPs | 224 / 112 |
| 15010.34 | Blender GPU | 2033.41 |
| 244 | Forza Horizon 5 | 68 |
| 323 | The Witcher 3 | 114 |
| 265 | Counter-Strike 2 | 108 |
| 202 | Far Cry 6 | 87 |
| 207 | Hogwarts Legacy | 50 |
| 327 | Call of Duty | 69 |
| 179 | Ghost of Tsushima | 51 |
| 229 | Cyberpunk 2077 | 69 |
| 299 | Tomb Raider | 98 |
| 253 | Average FPS | 79 |
| 294 | 1080p High | 98 |
| 253 | 1080p Ultra | 79 |
| 207 | 1440p Ultra | 59 |
| 124 | 4K Ultra | 33 |
| Low | Margin of Error | Low |
| 419309 | GB6 Compute Score | 99280 |
| 333 img/sec | Background Blur | 145.2 img/sec |
| 298.5 img/sec | Face Detection | 82 img/sec |
| 21.7 Gpixels/sec | Horizon Detection | 3.35 Gpixels/sec |
| 33.7 Gpixels/sec | Edge Detection | 5.01 Gpixels/sec |
| 34.5 Gpixels/sec | Gaussian Blur | 7.59 Gpixels/sec |
| 2.53 Gpixels/sec | Feature Matching | 0.81 Gpixels/sec |
| 2850 Gpixels/sec | Stereo Matching | 462.4 Gpixels/sec |
| 59146.9 FPS | Particle Physics | 18623.1 FPS |
| OpenCL | API | OpenCL |
| 38955 | G3D Mark Score | 12620 |
| 1413 | G2D Mark | 737 |
| 341 FPS | DirectX 11 | 73 FPS |
| 178 FPS | DirectX 12 | 70 FPS |
| 24632 Ops/s | GPU Compute | 5307 Ops/s |
| 114685 | Face Detection (HP) | 18556 |
| 344729 | Pose Estimation (Q) | 58801 |
| 4464 | Text Classification (SP) | 824 |
| 8184 | Machine Translation (HP) | 2397 |
| 27802 | Object Detection (SP) | 6895 |
| 85572 | Depth Estimation (Q) | 22219 |
| 687080 | Style Transfer (SP) | 161148 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 53861 | GB6 ML Single Precision | 12178 |
| 81080 | GB6 ML Half Precision | 12796 |
| 38938 | GB6 ML Quantized | 9086 |
| 18948 | Image Classification (SP) | 5475 |
| 55109 | Image Segmentation (HP) | 13781 |
| 58478 | Image Super Resolution (Q) | 11612 |
| 424 GPixel/s | Pixel Fill Rate | 269 GPixel/s |
| 1637 GTexel/s | Texture Fill Rate | 538 GTexel/s |
| 104.8 TFLOPS | FLOPS (FP32) | 17.2 TFLOPS |
| GDDR7 | Memory Type | GDDR6 |
| 32 GB | Memory Size | 8 GB |
| 1750 MHz | Memory Clock | 2000 MHz |
| 28000 Mbps | Effective Memory Speed | 16000 Mbps |
| 512-bit | Bus | 256-bit |
| No | ECC | No |
| 1792 GB/s | Memory Bandwidth | 512 GB/s |
| PCIe 5.0 x16 | Interface | PCIe 4.0 x16 |
| 575 W | TGP | 225 W |
| TSMC | Manufacturing | TSMC |
| 4 nm | Fabrication Process | 6 nm |
| 750 mm² | Die Size | 406 mm² |
| 92.2 billion | Transistor Count | 21 billion |
| 122.93 MTr/mm² | Transistor Density | 51.72 MTr/mm² |
| 12.2 | DirectX | 12 |
| 1.4 | Vulkan | 1.3 |
| 4.6 | OpenGL | 4.6 |
| 3 | OpenCL | 3 |
| 12 | CUDA | — |
| Yes | Ray Tracing | Yes |
| DLSS 4 | DLSS | No |
| 2.1b | DisplayPort | 2 |
| Nvidia | Vendor | Intel |
| Discrete | Build | Discrete |
| January 7, 2025 | Released | October 12, 2023 |
| Desktop | Case | Desktop |
| Gaming | Purpose | Gaming |
| High-end | Segment | Mid-range |
| Blackwell 2.0 | Architecture | Generation 12.7 |
| GB202-300 | GPU Codename | DG2-512 |
| -Intel Core Ultra 9 285Kor above | Recommended CPU | -Intel Core i5 12500or above |
| 52638 | Steel Nomad Lite Score | 10846 |
| 46941 | Time Spy | 12425 |
| 235224 | Solar Bay | 45771 |
| 38113 | Port Royal | 6594 |
| 87433 | Fire Strike | 30444 |
| 108736 | Wild Life Extreme | 23682 |
| 207058 | Night Raid | 110785 |
| 2017 MHz | Base Clock | 2050 MHz |
| 2407 MHz | Boost Clock | 2400 MHz |
| 21760 | Shading Units | 3584 |
| 680 | Texture Mapping Units (TMUs) | 224 |
| 176 | Render Output Units (ROPs) | 112 |
| 170 | Compute Units (Pipelines) | — |
| 680 | Tensor Cores | 448 |
| 170 | Ray-tracing Cores | 28 |
| 128KB per cluster | L1 Cache | — |
| 96MB shared | L2 Cache | 16MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare GeForce RTX 5090 with another →