GeForce RTX 3090
- Graphics Processor: GA102
- Cores: 10496
- TMUs / ROPs: 328 / 112
Review
Compare GeForce RTX 3090 with other GPUs
Games
| Forza Horizon 5 | 186 128 117 91 |
| The Witcher 3 | 205 170 128 66 |
| Counter-Strike 2 | 275 207 155 83 |
| Far Cry 6 | 156 145 124 77 |
| Hogwarts Legacy | 128 108 86 47 |
| CoD: Modern Warfare III | 171 159 123 80 |
| Ghost of Tsushima | 120 97 85 48 |
| Cyberpunk 2077 | 148 132 90 43 |
| Shadow of the Tomb Raider | 237 226 168 91 |
| 1080p High | 181 |
| 1080p Ultra | 152 |
| 1440p Ultra | 120 |
| 4K Ultra | 70 |
| Margin of Error | Low |
Test in Benchmarks
Performance tests in graphics benchmarks and applications3D Mark
| Time Spy | 19896 |
| Solar Bay | 97395 |
| Port Royal | 13633 |
| Fire Strike | 47512 |
| Wild Life Extreme | 43669 |
| Night Raid | 139766 |
GeekBench 6 OpenCL
| Background Blur | 320.8 img/sec |
| Face Detection | 207.2 img/sec |
| Horizon Detection | 10.1 Gpixels/sec |
| Edge Detection | 14.4 Gpixels/sec |
| Gaussian Blur | 11.5 Gpixels/sec |
| Feature Matching | 1.48 Gpixels/sec |
| Stereo Matching | 984.2 Gpixels/sec |
| Particle Physics | 27764.9 FPS |
| API | OpenCL |
Cinebench 2024 GPU
Passmark Graphics
| G2D Mark | 1069 |
| DirectX 11 | 218 FPS |
| DirectX 12 | 109 FPS |
| GPU Compute | 15028 Ops/s |
Artificial Intelligence Tests
Performance in machine learning and artificial intelligence tasksGeekBench 6 ML
| Image Classification (SP) | 10665 |
| Image Segmentation (HP) | 28963 |
| Image Super Resolution (Q) | 22697 |
| Face Detection (HP) | 55232 |
| Pose Estimation (Q) | 128155 |
| Text Classification (SP) | 3152 |
| Machine Translation (HP) | 5088 |
| Object Detection (SP) | 13891 |
| Depth Estimation (Q) | 37984 |
| Style Transfer (SP) | 272000 |
| Framework | ONNX |
| Backend | DirectML |
Specifications
Detailed technical specifications of the GeForce RTX 3090General
| Vendor | Nvidia |
| Build | Discrete |
| Released | September 24, 2020 |
| Launch price (MSRP) | $1499 |
| Case | Desktop |
| Purpose | Gaming |
| Segment | High-end |
| Architecture | Ampere |
| GPU Codename | GA102 |
| AMD Equivalent | - Radeon RX 6900 XT |
| Successor | - GeForce RTX 5090 |
| Recommended CPU | - Intel Core i9 12900K or above |
Desktop GPU rating - 20th place
Graphics Processing Unit
| Base Clock | 1395 MHz |
| Boost Clock | 1695 MHz |
| Shading Units | 10496 |
| Texture Mapping Units (TMUs) | 328 |
| Render Output Units (ROPs) | 112 |
| Compute Units (Pipelines) | 82 |
| Tensor Cores | 328 |
| Ray-tracing Cores | 82 |
| L1 Cache | 128KB per cluster |
| L2 Cache | 6MB shared |
| Instructions Per Cycle | 2 IPC |
Raw Performance
| Pixel Fill Rate | 190 GPixel/s |
| Texture Fill Rate | 556 GTexel/s |
Physical
| Interface | PCIe 4.0 x16 |
| TGP | 350 W |
| Manufacturing | Samsung |
| Fabrication Process | 8 nm |
| Die Size | 628 mm² |
| Transistor Count | 28 billion |
| Transistor Density | 44.59 MTr/mm² |
Memory
| Memory Type | GDDR6X |
| Memory Size | 24 GB |
| Memory Clock | 1219 MHz |
| Effective Memory Speed | 19500 Mbps |
| Bus | 384-bit |
| ECC | No |
API
| DirectX | 12 |
| Vulkan | 1.3 |
| OpenGL | 4.6 |
| OpenCL | 3.0 |
| CUDA | 8.6 |
| Ray Tracing | Yes |
| DLSS | DLSS 2 |
| DisplayPort | 1.4a |