Quick/Comparisons
66 | Gaming | 5 |
56 | Workstation | 13 |
63 | AI/ML |
Compare with something else?
Compare Radeon RX 9070 with another →6 |
49 | Energy Efficiency | 75 |
62 | Quick Comparison Final Score | 19 |
| Navi 48 | Graphics Processor | — |
| 3584 | Cores | 448 |
| 224 / 128 | TMUs / ROPs | 28 / 8 |
| 2749.68 | Blender GPU | — |
| 156 | Forza Horizon 5 | 10 |
| 217 | The Witcher 3 | — |
| 218 | Counter-Strike 2 | 14 |
| 164 | Far Cry 6 | 11 |
| 123 | Hogwarts Legacy | - |
| 206 | Call of Duty | 13 |
| 123 | Ghost of Tsushima | - |
| 149 | Cyberpunk 2077 | - |
| 240 | Tomb Raider | 11 |
| 177 | Average FPS | 10 |
| 213 | 1080p High | 13 |
| 177 | 1080p Ultra | 10 |
| 139 | 1440p Ultra | - |
| 82 | 4K Ultra | - |
| Low | Margin of Error | Low |
| 149949 | GB6 Compute Score | 14408 |
| 377.5 img/sec | Background Blur | 33.7 img/sec |
| 221.3 img/sec | Face Detection | 15.4 img/sec |
| 6.4 Gpixels/sec | Horizon Detection | 0.35 Gpixels/sec |
| 8.99 Gpixels/sec | Edge Detection | 0.76 Gpixels/sec |
| 8.52 Gpixels/sec | Gaussian Blur | 0.69 Gpixels/sec |
| 1.36 Gpixels/sec | Feature Matching | 0.21 Gpixels/sec |
| 606 Gpixels/sec | Stereo Matching | 54.7 Gpixels/sec |
| 8521.2 FPS | Particle Physics | 1957.6 FPS |
| OpenCL | API | OpenCL |
| 25389 | G3D Mark Score | 2250 |
| 1265 | G2D Mark | 568 |
| 280 FPS | DirectX 11 | 16 FPS |
| 74 FPS | DirectX 12 | 11 FPS |
| 14791 Ops/s | GPU Compute | 1215 Ops/s |
| 37056 | GB6 ML Single Precision | 3157 |
| 54866 | GB6 ML Half Precision | 4191 |
| 29946 | GB6 ML Quantized | 2365 |
| 15375 | Image Classification (SP) | 1323 |
| 37834 | Image Segmentation (HP) | 1675 |
| 43910 | Image Super Resolution (Q) | 2919 |
| 67858 | Face Detection (HP) | 4386 |
| 172939 | Pose Estimation (Q) | 9348 |
| 4469 | Text Classification (SP) | 1024 |
| 8499 | Machine Translation (HP) | 1888 |
| 15393 | Object Detection (SP) | 1452 |
| 66890 | Depth Estimation (Q) | 5272 |
| 365633 | Style Transfer (SP) | 19585 |
| ONNX | Framework | ONNX |
| DirectML | Backend | DirectML |
| 323 GPixel/s | Pixel Fill Rate | 12-16 GPixel/s |
| 564 GTexel/s | Texture Fill Rate | 42-56 GTexel/s |
| 36.1 TFLOPS | FLOPS (FP32) | 1.3-1.8 TFLOPS |
| GDDR6 | Memory Type | System Shared |
| 16 GB | Memory Size | System RAM Dependent |
| 2518 MHz | Memory Clock | System RAM Dependent |
| 20100 Mbps | Effective Memory Speed | System RAM Dependent |
| 256-bit | Bus | — |
| No | ECC | No |
| 640 GB/s | Memory Bandwidth | — |
| PCIe 5.0 x16 | Interface | IGP |
| 220 W | TGP | 10-45 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 4 nm | Fabrication Process | 7 nm |
| 356.5 mm² | Die Size | — |
| 53.9 billion | Transistor Count | — |
| 151.19 MTr/mm² | Transistor Density | — |
| 12 | DirectX | 12.1 |
| 1.3 | Vulkan | — |
| 4.6 | OpenGL | 4.6 |
| 2.2 | OpenCL | 2.2 |
| No | CUDA | — |
| Yes | Ray Tracing | No |
| No | DLSS | No |
| 2.1a | DisplayPort | — |
| Amd | Vendor | Amd |
| Discrete | Build | Integrated |
| January 7, 2025 | Released | January 7, 2020 |
| Desktop | Case | Laptop |
| Gaming | Purpose | Generic |
| High-end | Segment | Entry-level |
| RDNA 4.0 | Architecture | Raven |
| Navi 48 | GPU Codename | — |
| -Intel Core Ultra 5 245Kor above | Recommended CPU | -AMD Ryzen 5 5500Uor above |
| 22783 | Steel Nomad Lite Score | 943 |
| 26826 | Time Spy | 1068 |
| 105512 | Solar Bay | — |
| 16089 | Port Royal | — |
| 60226 | Fire Strike | 3396 |
| 48184 | Wild Life Extreme | 2015 |
| 176216 | Night Raid | 13093 |
| 1330 MHz | Base Clock | 300 MHz |
| 2520 MHz | Boost Clock | 1500-2000 MHz (configurable) |
| 3584 | Shading Units | 448 |
| 224 | Texture Mapping Units (TMUs) | 28 |
| 128 | Render Output Units (ROPs) | 8 |
| 56 | Compute Units (Pipelines) | 7 |
| 112 | Tensor Cores | No |
| 56 | Ray-tracing Cores | No |
| 128KB per cluster | L1 Cache | — |
| 4MB shared | L2 Cache | — |
| 4 IPC | Instructions Per Cycle | 2 IPC |