Quick/Comparisons
15 | Gaming | 28 |
25 | Workstation | 31 |
21 | AI/ML |
Compare with something else?
Compare Apple M4 GPU with another →28 |
95 | Energy Efficiency | 72 |
32 | Quick Comparison Final Score | 36 |
| Custom | Graphics Processor | Navi 33 |
| 1280 | Cores | 1792 |
| 80 / 40 | TMUs / ROPs | 112 / 64 |
| 1062.09 | Blender GPU | 983.46 |
| — | Tomb Raider | 102 |
| — | Average FPS | 78 |
| — | Forza Horizon 5 | 69 |
| — | The Witcher 3 | 90 |
| — | Counter-Strike 2 | 122 |
| — | Far Cry 6 | 84 |
| — | Hogwarts Legacy | 52 |
| — | Call of Duty | 77 |
| — | Ghost of Tsushima | 54 |
| — | Cyberpunk 2077 | 56 |
| — | 1080p High | 101 |
| — | 1080p Ultra | 78 |
| — | 1440p Ultra | 55 |
| — | 4K Ultra | 30 |
| — | Margin of Error | Medium |
| 37678 | GB6 Compute Score | 70637 |
| 72.2 img/sec | Background Blur | 171.3 img/sec |
| 48.3 img/sec | Face Detection | 88.4 img/sec |
| 1.4 Gpixels/sec | Horizon Detection | 2.56 Gpixels/sec |
| 1.94 Gpixels/sec | Edge Detection | 3.44 Gpixels/sec |
| 1.5 Gpixels/sec | Gaussian Blur | 3.12 Gpixels/sec |
| 0.53 Gpixels/sec | Feature Matching | 0.85 Gpixels/sec |
| 123.3 Gpixels/sec | Stereo Matching | 225.1 Gpixels/sec |
| 4938.4 FPS | Particle Physics | 8759.4 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 15558 |
| — | G2D Mark | 769 |
| — | DirectX 11 | 140 FPS |
| — | DirectX 12 | 66 FPS |
| — | GPU Compute | 6568 Ops/s |
| 10346 | GB6 ML Single Precision | 15936 |
| 11883 | GB6 ML Half Precision | 19176 |
| 11036 | GB6 ML Quantized | 13239 |
| 4874 | Image Classification (SP) | 7912 |
| 8168 | Image Segmentation (HP) | 8340 |
| 11983 | Image Super Resolution (Q) | 19913 |
| 17036 | Face Detection (HP) | 25456 |
| 31171 | Pose Estimation (Q) | 77820 |
| 2964 | Text Classification (SP) | 3009 |
| 3410 | Machine Translation (HP) | 2738 |
| 5112 | Object Detection (SP) | 6115 |
| 22700 | Depth Estimation (Q) | 34757 |
| 60872 | Style Transfer (SP) | 148255 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 59 GPixel/s | Pixel Fill Rate | 100-119 GPixel/s |
| 118 GTexel/s | Texture Fill Rate | 175-209 GTexel/s |
| 3.8 TFLOPS | FLOPS (FP32) | 11.2-13.4 TFLOPS |
| System Shared | Memory Type | GDDR6 |
| — | Memory Size | 8 GB |
| 7500 MHz | Memory Clock | 2000 MHz |
| 15000 Mbps | Effective Memory Speed | 16000 Mbps |
| 128-bit | Bus | 128-bit |
| No | ECC | No |
| 120 GB/s | Memory Bandwidth | 256 GB/s |
| Custom | Interface | PCIe 4.0 x8 |
| 18 W | TGP | 50-75 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 6 nm |
| — | Die Size | 204 mm² |
| 28 billion | Transistor Count | 13.3 billion |
| — | Transistor Density | 65.2 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 2.2 |
| — | CUDA | — |
| Yes | Ray Tracing | Yes |
| No | DLSS | No |
| — | DisplayPort | 2.1 |
| Apple | Vendor | Amd |
| Integrated | Build | Discrete |
| May 7, 2024 | Released | January 4, 2023 |
| Laptop | Case | Laptop |
| Professional | Purpose | Gaming |
| Mid-range | Segment | Mid-range |
| Apple M GPU | Architecture | RDNA 3.0 |
| Custom | GPU Codename | Navi 33 |
| -Apple M4 (10-Core)or above | Recommended CPU | -AMD Ryzen 7 7735HSor above |
| 4097 | Steel Nomad Lite Score | 7954 |
| — | Time Spy | 9185 |
| 16545 | Solar Bay | 35056 |
| — | Port Royal | 4692 |
| — | Fire Strike | 25593 |
| 9591 | Wild Life Extreme | 16053 |
| — | Night Raid | 95187 |
| 500 MHz | Base Clock | 1500 MHz |
| 1470 MHz | Boost Clock | 1560-1865 MHz (configurable) |
| 1280 | Shading Units | 1792 |
| 80 | Texture Mapping Units (TMUs) | 112 |
| 40 | Render Output Units (ROPs) | 64 |
| 160 | Compute Units (Pipelines) | 28 |
| — | Tensor Cores | — |
| — | Ray-tracing Cores | 28 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 4 IPC |