Technical Specifications
Performance
| Architecture | RDNA 3.5 |
|---|---|
| GPU Codename | Strix Halo |
| Base Clock | 1295 MHz |
| Boost Clock | 2900 MHz |
| Shading Units | 2560 |
| Ray-tracing Cores | 40 |
| Compute Units (Pipelines) | 40 |
Memory
| Memory Type | System Shared |
|---|
Graphics APIs
| DirectX | 12 |
|---|---|
| Vulkan | 1.4 |
| OpenGL | 4.6 |
| OpenCL | 2.1 |
| CUDA | No |
| Ray Tracing | Yes |
| DLSS | No |
| DisplayPort | 2.1a |
Technical
| TGP | 45-120 W (configurable) |
|---|---|
| Fabrication Process | 4 nm |
| Die Size | 308 mm² |
| Transistor Count | 34 billion |
| Interface | PCIe 5.0 x16 |
| Max. Temperature | 100°C |
| Manufacturing | TSMC |
| Vendor | Amd |
| Build | Integrated |
| Released | January 7, 2025 |
Other Specifications
| TGP / Version | 80W (Whole SoC TDP)70W (Whole SoC TDP)60W (Whole SoC TDP)50W (Whole SoC TDP)40W (Whole SoC TDP)30W (Whole SoC TDP)15W (Whole SoC TDP) |
|---|---|
| Forza Horizon 5 | 138766541 |
| The Witcher 3 | 1251047239 |
| Counter-Strike 2 | 1971359450 |
| Far Cry 6 | 107956939 |
| Hogwarts Legacy | 74594324 |
| CoD: Modern Warfare III | 99906438 |
| Ghost of Tsushima | 78624724 |
| Cyberpunk 2077 | 77654117 |
| Shadow of the Tomb Raider | 1321157940 |
| 1080p High | 114 |
| 1080p Ultra | 89 |
| 1440p Ultra | 64 |
| 4K Ultra | 35 |
| Margin of Error | Low |
| Time Spy | 9930 |
| Solar Bay | 47136 |
| Port Royal | 5466 |
| Fire Strike | 27561 |
| Wild Life Extreme | 18951 |
| Night Raid | 105526 |
| Background Blur | 232.1 img/sec |
| Face Detection | 122.5 img/sec |
| Horizon Detection | 2.92 Gpixels/sec |
| Edge Detection | 4.07 Gpixels/sec |
| Gaussian Blur | 3.49 Gpixels/sec |
| Feature Matching | 1.08 Gpixels/sec |
| Stereo Matching | 396.3 Gpixels/sec |
| Particle Physics | 13016.5 FPS |
| G2D Mark | 1177 |
| DirectX 11 | 157 FPS |
| DirectX 12 | 84 FPS |
| GPU Compute | 9061 Ops/s |
| Image Classification (SP) | 11176 |
| Image Segmentation (HP) | 14027 |
| Image Super Resolution (Q) | 27198 |
| Face Detection (HP) | 40613 |
| Pose Estimation (Q) | 92878 |
| Text Classification (SP) | 4158 |
| Machine Translation (HP) | 6473 |
| Object Detection (SP) | 7431 |
| Depth Estimation (Q) | 41825 |
| Style Transfer (SP) | 190221 |
| Framework | ONNX |
| Backend | DirectML |
| Case | Laptop |
| Purpose | Generic |
| Segment | Mid-range |
| Rival Equivalent | GeForce RTX 4060 Laptop |
| Texture Mapping Units (TMUs) | 160 |
| Render Output Units (ROPs) | 64 |
| L1 Cache | 128KB per cluster |
| L2 Cache | 8MB shared |
| Instructions Per Cycle | 2 IPC |
| Pixel Fill Rate | 186 GPixel/s |
| Texture Fill Rate | 464 GTexel/s |
| Transistor Density | 110.39 MTr/mm² |