Quick/Comparisons
14 | Gaming | 9 |
24 | Workstation | 12 |
| — | AI/ML |
Compare with something else?
Compare Apple M4 GPU with another →97 | Energy Efficiency | 93 |
32 | Quick Comparison Final Score | 24 |
| Custom | Graphics Processor | Phoenix |
| 1024 | Cores | 512 |
| 64 / 32 | TMUs / ROPs | 32 / 16 |
| 1060.26 | Blender GPU | 258.41 |
| — | Forza Horizon 5 | 20 |
| — | The Witcher 3 | 20 |
| — | Counter-Strike 2 | 26 |
| — | Far Cry 6 | 24 |
| — | Hogwarts Legacy | 11 |
| — | Call of Duty | 26 |
| — | Ghost of Tsushima | 12 |
| — | Cyberpunk 2077 | 16 |
| — | Tomb Raider | 21 |
| — | Average FPS | 20 |
| — | 1080p High | 25 |
| — | 1080p Ultra | 20 |
| — | 1440p Ultra | 13 |
| — | 4K Ultra | - |
| — | Margin of Error | Low |
| 30791 | GB6 Compute Score | 22589 |
| 57.4 img/sec | Background Blur | 57.8 img/sec |
| 41.3 img/sec | Face Detection | 28.7 img/sec |
| 1.21 Gpixels/sec | Horizon Detection | 0.8 Gpixels/sec |
| 1.74 Gpixels/sec | Edge Detection | 1 Gpixels/sec |
| 1.12 Gpixels/sec | Gaussian Blur | 0.82 Gpixels/sec |
| 0.43 Gpixels/sec | Feature Matching | 0.29 Gpixels/sec |
| 94.9 Gpixels/sec | Stereo Matching | 85.4 Gpixels/sec |
| 3902.5 FPS | Particle Physics | 2844.5 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 5374 |
| — | G2D Mark | 894 |
| — | DirectX 11 | 53 FPS |
| — | DirectX 12 | 25 FPS |
| — | GPU Compute | 2877 Ops/s |
| — | GB6 ML Single Precision | 8097 |
| — | GB6 ML Half Precision | 11246 |
| — | GB6 ML Quantized | 6004 |
| — | Image Classification (SP) | 3449 |
| — | Image Segmentation (HP) | 4098 |
| — | Image Super Resolution (Q) | 7937 |
| — | Face Detection (HP) | 12168 |
| — | Pose Estimation (Q) | 29915 |
| — | Text Classification (SP) | 2571 |
| — | Machine Translation (HP) | 3852 |
| — | Object Detection (SP) | 2687 |
| — | Depth Estimation (Q) | 14163 |
| — | Style Transfer (SP) | 59966 |
| — | Framework | ONNX |
| — | Backend | DirectML |
| 47 GPixel/s | Pixel Fill Rate | 42-45 GPixel/s |
| 94 GTexel/s | Texture Fill Rate | 83-90 GTexel/s |
| 3 TFLOPS | FLOPS (FP32) | 2.7-2.9 TFLOPS |
| System Shared | Memory Type | System Shared |
| — | Memory Size | — |
| 7500 MHz | Memory Clock | — |
| 15000 Mbps | Effective Memory Speed | — |
| 128-bit | Bus | — |
| No | ECC | — |
| 120 GB/s | Memory Bandwidth | — |
| Custom | Interface | PCIe 4.0 x8 |
| 15 W | TGP | 15 W |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 4 nm |
| — | Die Size | 178 mm² |
| — | Transistor Count | 25.4 billion |
| — | Transistor Density | 142.7 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 2.1 |
| — | CUDA | — |
| Yes | Ray Tracing | Yes |
| No | DLSS | No |
| — | DisplayPort | 2.1 |
| Apple | Vendor | Amd |
| Integrated | Build | Integrated |
| October 29, 2024 | Released | January 4, 2023 |
| Laptop | Case | Laptop |
| Professional | Purpose | Generic |
| Mid-range | Segment | Entry-level |
| Apple M GPU | Architecture | RDNA 3.0 |
| Custom | GPU Codename | Phoenix |
| -Apple M4 (8-Core)or above | Recommended CPU | -AMD Ryzen 5 7640Uor above |
| 3450 | Steel Nomad Lite Score | 2061 |
| — | Time Spy | 2373 |
| 13908 | Solar Bay | 9378 |
| — | Port Royal | 975 |
| — | Fire Strike | 7087 |
| 7928 | Wild Life Extreme | 4630 |
| — | Night Raid | 29647 |
| 500 MHz | Base Clock | 800 MHz |
| 1470 MHz | Boost Clock | 2600-2800 MHz (configurable) |
| 1024 | Shading Units | 512 |
| 64 | Texture Mapping Units (TMUs) | 32 |
| 32 | Render Output Units (ROPs) | 16 |
| 128 | Compute Units (Pipelines) | 8 |
| — | Tensor Cores | — |
| — | Ray-tracing Cores | 6 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |