Quick/Comparisons
14 | Gaming | 5 |
23 | Workstation | 11 |
10 | AI/ML |
Compare with something else?
Compare Apple M3 GPU with another →4 |
95 | Energy Efficiency | 89 |
30 | Quick Comparison Final Score | 20 |
| — | Graphics Processor | Mendocino |
| 1280 | Cores | 128 |
| 80 / 40 | TMUs / ROPs | 8 / 4 |
| 155.27 | Blender GPU | 47.24 |
| — | Forza Horizon 5 | — |
| — | The Witcher 3 | — |
| — | Counter-Strike 2 | — |
| — | Far Cry 6 | — |
| — | Hogwarts Legacy | — |
| — | Call of Duty | — |
| — | Ghost of Tsushima | — |
| — | Cyberpunk 2077 | — |
| — | Tomb Raider | — |
| — | Average FPS | — |
| — | 1080p High | — |
| — | 1080p Ultra | — |
| — | 1440p Ultra | — |
| — | 4K Ultra | — |
| — | Margin of Error | — |
| 30027 | GB6 Compute Score | 5099 |
| 60.3 img/sec | Background Blur | 12 img/sec |
| 40.8 img/sec | Face Detection | 6.42 img/sec |
| 1.21 Gpixels/sec | Horizon Detection | 0.16 Gpixels/sec |
| 1.64 Gpixels/sec | Edge Detection | 0.23 Gpixels/sec |
| 1.2 Gpixels/sec | Gaussian Blur | 0.16 Gpixels/sec |
| 0.37 Gpixels/sec | Feature Matching | 0.07 Gpixels/sec |
| 102.6 Gpixels/sec | Stereo Matching | 18.2 Gpixels/sec |
| 3293.3 FPS | Particle Physics | 828.6 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 1810 |
| — | G2D Mark | 536 |
| — | DirectX 11 | 17 FPS |
| — | DirectX 12 | 7 FPS |
| — | GPU Compute | 771 Ops/s |
| 4120 | GB6 ML Single Precision | 1917 |
| 6992 | GB6 ML Half Precision | 2053 |
| 5638 | GB6 ML Quantized | 1571 |
| 1579 | Image Classification (SP) | 660 |
| 2944 | Image Segmentation (HP) | 627 |
| 4425 | Image Super Resolution (Q) | 2053 |
| 6661 | Face Detection (HP) | 2153 |
| 15020 | Pose Estimation (Q) | 5787 |
| 2984 | Text Classification (SP) | 713 |
| 5876 | Machine Translation (HP) | 1114 |
| 1826 | Object Detection (SP) | 709 |
| 9339 | Depth Estimation (Q) | 3233 |
| 22291 | Style Transfer (SP) | 11091 |
| Core ML | Framework | ONNX |
| CPU | Backend | DirectML |
| 55 GPixel/s | Pixel Fill Rate | 8-9 GPixel/s |
| 110 GTexel/s | Texture Fill Rate | 15-18 GTexel/s |
| 3.5 TFLOPS | FLOPS (FP32) | 0.5-0.6 TFLOPS |
| System Shared | Memory Type | System Shared |
| — | Memory Size | — |
| 6400 MHz | Memory Clock | — |
| 12800 Mbps | Effective Memory Speed | — |
| 128-bit | Bus | — |
| No | ECC | No |
| 102.4 GB/s | Memory Bandwidth | — |
| Custom | Interface | PCIe 4.0 x8 |
| 15 W | TGP | 15 W |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 6 nm |
| — | Die Size | 122 mm² |
| 13.8 billion | Transistor Count | 3.4 billion |
| — | Transistor Density | 27.87 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 2 |
| — | CUDA | — |
| Yes | Ray Tracing | Yes |
| No | DLSS | No |
| — | DisplayPort | — |
| Apple | Vendor | Amd |
| Integrated | Build | Integrated |
| October 31, 2023 | Released | October 1, 2022 |
| Laptop | Case | Laptop |
| Professional | Purpose | Generic |
| Mid-range | Segment | Entry-level |
| Apple M GPU | Architecture | RDNA 2.0 |
| — | GPU Codename | Mendocino |
| -Apple M3or above | Recommended CPU | -AMD Ryzen 5 7520Uor above |
| 3478 | Steel Nomad Lite Score | 447 |
| — | Time Spy | 519 |
| 13529 | Solar Bay | 2078 |
| — | Port Royal | 309 |
| — | Fire Strike | 1760 |
| 8229 | Wild Life Extreme | 994 |
| — | Night Raid | 7447 |
| 500 MHz | Base Clock | 400 MHz |
| 1380 MHz | Boost Clock | 1900-2200 MHz (configurable) |
| 1280 | Shading Units | 128 |
| 80 | Texture Mapping Units (TMUs) | 8 |
| 40 | Render Output Units (ROPs) | 4 |
| 160 | Compute Units (Pipelines) | 2 |
| — | Tensor Cores | — |
| — | Ray-tracing Cores | 2 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |