Technical Specifications
Performance
| Architecture | RDNA 3.0 |
|---|---|
| GPU Codename | Phoenix |
| Base Clock | 800 MHz |
| Boost Clock | 2700-2800 MHz (configurable) |
| Shading Units | 768 |
| Ray-tracing Cores | 12 |
| Compute Units (Pipelines) | 12 |
Memory
| Memory Type | System Shared |
|---|
Graphics APIs
| DirectX | 12 |
|---|---|
| Vulkan | 1.3 |
| OpenGL | 4.6 |
| OpenCL | 2.1 |
| Ray Tracing | Yes |
| DLSS | No |
| DisplayPort | 2.1 |
Technical
| TGP | 15 W |
|---|---|
| Fabrication Process | 4 nm |
| Die Size | 178 mm² |
| Transistor Count | 25.4 billion |
| Interface | PCIe 4.0 x8 |
| Manufacturing | TSMC |
| Vendor | Amd |
| Build | Integrated |
| Released | January 4, 2023 |
Other Specifications
| Forza Horizon 5 | 52262211 |
|---|---|
| The Witcher 3 | 262613 - |
| Counter-Strike 2 | 4431219 |
| Far Cry 6 | 39292110 |
| Hogwarts Legacy | 221511 - |
| CoD: Modern Warfare III | 34291912 |
| Ghost of Tsushima | 211510 - |
| Cyberpunk 2077 | 192010 - |
| Shadow of the Tomb Raider | 3631209 |
| 1080p High | 33 |
| 1080p Ultra | 25 |
| 1440p Ultra | 16 |
| Margin of Error | Low |
| Time Spy | 2786 |
| Solar Bay | 11359 |
| Port Royal | 1585 |
| Fire Strike | 7835 |
| Wild Life Extreme | 5269 |
| Night Raid | 33674 |
| Background Blur | 86.4 img/sec |
| Face Detection | 35.8 img/sec |
| Horizon Detection | 1.03 Gpixels/sec |
| Edge Detection | 1.36 Gpixels/sec |
| Gaussian Blur | 1.41 Gpixels/sec |
| Feature Matching | 0.41 Gpixels/sec |
| Stereo Matching | 132.4 Gpixels/sec |
| Particle Physics | 4470.7 FPS |
| API | OpenGL |
| G2D Mark | 890 |
| DirectX 11 | 66 FPS |
| DirectX 12 | 30 FPS |
| GPU Compute | 3658 Ops/s |
| Aztec Ruins Normal Tier | 183.4 FPS |
| Car Chase | 205.6 FPS |
| Manhattan 3.1.1 (1440p) | 161.6 FPS |
| T-Rex | 713.2 FPS |
| Tessellation | 1373.8 FPS |
| ALU 2 | 875 FPS |
| Driver Overhead 2 | 268.9 FPS |
| OS | Windows |
| Image Classification (SP) | 3860 |
| Image Segmentation (HP) | 5435 |
| Image Super Resolution (Q) | 10580 |
| Face Detection (HP) | 12855 |
| Pose Estimation (Q) | 34716 |
| Text Classification (SP) | 2578 |
| Machine Translation (HP) | 3897 |
| Object Detection (SP) | 3078 |
| Depth Estimation (Q) | 14989 |
| Style Transfer (SP) | 78608 |
| Framework | ONNX |
| Backend | DirectML |
| Case | Laptop |
| Purpose | Generic |
| Segment | Mid-range |
| Rival Equivalent | Intel Arc iGPU (8-Cores) |
| Texture Mapping Units (TMUs) | 48 |
| Render Output Units (ROPs) | 32 |
| L1 Cache | 128KB per cluster |
| L2 Cache | 2MB shared |
| Instructions Per Cycle | 2 IPC |
| Pixel Fill Rate | 86-90 GPixel/s |
| Texture Fill Rate | 130-134 GTexel/s |
| Transistor Density | 142.7 MTr/mm² |