Quick/Comparisons
14 | Gaming | 14 |
24 | Workstation | 24 |
| — | AI/ML | 14 |
Compare with something else?
Compare Apple M4 GPU with another →97 | Energy Efficiency | 24 |
32 | Quick Comparison Final Score | 17 |
| Custom | Graphics Processor | Polaris 20 |
| 1024 | Cores | 2304 |
| 64 / 32 | TMUs / ROPs | 144 / 32 |
| 1060.26 | Blender GPU | 391.86 |
| — | Forza Horizon 5 | 38 |
| — | The Witcher 3 | 41 |
| — | Counter-Strike 2 | 56 |
| — | Far Cry 6 | 46 |
| — | Hogwarts Legacy | 28 |
| — | Call of Duty | 41 |
| — | Ghost of Tsushima | 24 |
| — | Cyberpunk 2077 | 28 |
| — | Tomb Raider | 52 |
| — | Average FPS | 39 |
| — | 1080p High | 51 |
| — | 1080p Ultra | 39 |
| — | 1440p Ultra | 26 |
| — | 4K Ultra | 13 |
| — | Margin of Error | Low |
| 30791 | GB6 Compute Score | 40514 |
| 57.4 img/sec | Background Blur | 59.5 img/sec |
| 41.3 img/sec | Face Detection | 33.8 img/sec |
| 1.21 Gpixels/sec | Horizon Detection | 1.56 Gpixels/sec |
| 1.74 Gpixels/sec | Edge Detection | 3.1 Gpixels/sec |
| 1.12 Gpixels/sec | Gaussian Blur | 2.92 Gpixels/sec |
| 0.43 Gpixels/sec | Feature Matching | 0.33 Gpixels/sec |
| 94.9 Gpixels/sec | Stereo Matching | 161.8 Gpixels/sec |
| 3902.5 FPS | Particle Physics | 5395.1 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 8792 |
| — | G2D Mark | 769 |
| — | DirectX 11 | 60 FPS |
| — | DirectX 12 | 43 FPS |
| — | GPU Compute | 3431 Ops/s |
| — | GB6 ML Single Precision | 7905 |
| — | GB6 ML Half Precision | 8166 |
| — | GB6 ML Quantized | 5642 |
| — | Image Classification (SP) | 3143 |
| — | Image Segmentation (HP) | 3955 |
| — | Image Super Resolution (Q) | 7872 |
| — | Face Detection (HP) | 11922 |
| — | Pose Estimation (Q) | 41162 |
| — | Text Classification (SP) | 1199 |
| — | Machine Translation (HP) | 1773 |
| — | Object Detection (SP) | 3620 |
| — | Depth Estimation (Q) | 13995 |
| — | Style Transfer (SP) | 84083 |
| — | Framework | ONNX |
| — | Backend | DirectML |
| 47 GPixel/s | Pixel Fill Rate | 43 GPixel/s |
| 94 GTexel/s | Texture Fill Rate | 193 GTexel/s |
| 3 TFLOPS | FLOPS (FP32) | 6.2 TFLOPS |
| System Shared | Memory Type | GDDR5 |
| — | Memory Size | 8 GB |
| 7500 MHz | Memory Clock | 2000 MHz |
| 15000 Mbps | Effective Memory Speed | 8000 Mbps |
| 128-bit | Bus | 256-bit |
| No | ECC | No |
| 120 GB/s | Memory Bandwidth | 256 GB/s |
| Custom | Interface | PCIe 3.0 x16 |
| 15 W | TGP | 185 W |
| TSMC | Manufacturing | GlobalFoundries |
| 3 nm | Fabrication Process | 14 nm |
| — | Die Size | 232 mm² |
| — | Transistor Count | 5 billion |
| — | Transistor Density | 21.55 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 2.1 |
| — | CUDA | — |
| Yes | Ray Tracing | No |
| No | DLSS | No |
| — | DisplayPort | 1.4a |
| Apple | Vendor | Amd |
| Integrated | Build | Discrete |
| October 29, 2024 | Released | April 18, 2017 |
| Laptop | Case | Desktop |
| Professional | Purpose | Gaming |
| Mid-range | Segment | High-end |
| Apple M GPU | Architecture | GCN 4.0 |
| Custom | GPU Codename | Polaris 20 |
| -Apple M4 (8-Core)or above | Recommended CPU | -Intel Core i5 8500or above |
| 3450 | Steel Nomad Lite Score | 4182 |
| — | Time Spy | 4354 |
| 13908 | Solar Bay | — |
| — | Port Royal | 527 |
| — | Fire Strike | 14979 |
| 7928 | Wild Life Extreme | 8920 |
| — | Night Raid | 56872 |
| 500 MHz | Base Clock | 1257 MHz |
| 1470 MHz | Boost Clock | 1340 MHz |
| 1024 | Shading Units | 2304 |
| 64 | Texture Mapping Units (TMUs) | 144 |
| 32 | Render Output Units (ROPs) | 32 |
| 128 | Compute Units (Pipelines) | — |
| — | Tensor Cores | No |
| — | Ray-tracing Cores | No |
| — | L1 Cache | 16KB per cluster |
| — | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |