Quick/Comparisons
14 | Gaming | 37 |
23 | Workstation | 41 |
10 | AI/ML |
Compare with something else?
Compare Apple M3 GPU with another →31 |
95 | Energy Efficiency | 47 |
30 | Quick Comparison Final Score | 39 |
| — | Graphics Processor | GA104 (GN20-E6) |
| 1280 | Cores | 5632 |
| 80 / 40 | TMUs / ROPs | 176 / 80 |
| 155.27 | Blender GPU | 3120.06 |
| — | Forza Horizon 5 | 85 |
| — | The Witcher 3 | 123 |
| — | Counter-Strike 2 | 167 |
| — | Far Cry 6 | 101 |
| — | Hogwarts Legacy | 67 |
| — | Call of Duty | 96 |
| — | Ghost of Tsushima | 63 |
| — | Cyberpunk 2077 | 81 |
| — | Tomb Raider | 126 |
| — | Average FPS | 101 |
| — | 1080p High | 122 |
| — | 1080p Ultra | 101 |
| — | 1440p Ultra | 74 |
| — | 4K Ultra | 42 |
| — | Margin of Error | Low |
| 30027 | GB6 Compute Score | 116337 |
| 60.3 img/sec | Background Blur | 173.8 img/sec |
| 40.8 img/sec | Face Detection | 96.9 img/sec |
| 1.21 Gpixels/sec | Horizon Detection | 5.69 Gpixels/sec |
| 1.64 Gpixels/sec | Edge Detection | 7.75 Gpixels/sec |
| 1.2 Gpixels/sec | Gaussian Blur | 6.27 Gpixels/sec |
| 0.37 Gpixels/sec | Feature Matching | 1.14 Gpixels/sec |
| 102.6 Gpixels/sec | Stereo Matching | 480.4 Gpixels/sec |
| 3293.3 FPS | Particle Physics | 14706.2 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 17472 |
| — | G2D Mark | 773 |
| — | DirectX 11 | 152 FPS |
| — | DirectX 12 | 72 FPS |
| — | GPU Compute | 7362 Ops/s |
| 4120 | GB6 ML Single Precision | 17250 |
| 6992 | GB6 ML Half Precision | 27936 |
| 5638 | GB6 ML Quantized | 11249 |
| 1579 | Image Classification (SP) | 6969 |
| 2944 | Image Segmentation (HP) | 17142 |
| 4425 | Image Super Resolution (Q) | 16426 |
| 6661 | Face Detection (HP) | 31628 |
| 15020 | Pose Estimation (Q) | 74413 |
| 2984 | Text Classification (SP) | 2522 |
| 5876 | Machine Translation (HP) | 3639 |
| 1826 | Object Detection (SP) | 9002 |
| 9339 | Depth Estimation (Q) | 27790 |
| 22291 | Style Transfer (SP) | 160005 |
| Core ML | Framework | ONNX |
| CPU | Backend | DirectML |
| 55 GPixel/s | Pixel Fill Rate | 83-119 GPixel/s |
| 110 GTexel/s | Texture Fill Rate | 182-261 GTexel/s |
| 3.5 TFLOPS | FLOPS (FP32) | 11.7-16.7 TFLOPS |
| System Shared | Memory Type | GDDR6 |
| — | Memory Size | 8 GB |
| 6400 MHz | Memory Clock | 1750 MHz |
| 12800 Mbps | Effective Memory Speed | 14000 Mbps |
| 128-bit | Bus | 256-bit |
| No | ECC | No |
| 102.4 GB/s | Memory Bandwidth | 448 GB/s |
| Custom | Interface | PCIe 4.0 x16 |
| 15 W | TGP | 80-150 W (configurable) |
| TSMC | Manufacturing | Samsung |
| 3 nm | Fabrication Process | 8 nm |
| — | Die Size | 392 mm² |
| 13.8 billion | Transistor Count | 17.4 billion |
| — | Transistor Density | 44.39 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.2 |
| — | OpenGL | 4.6 |
| — | OpenCL | 3 |
| — | CUDA | 8.6 |
| Yes | Ray Tracing | Yes |
| No | DLSS | DLSS 2 |
| — | DisplayPort | 1.4a |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| October 31, 2023 | Released | — |
| Laptop | Case | Laptop |
| Professional | Purpose | Gaming |
| Mid-range | Segment | High-end |
| Apple M GPU | Architecture | Ampere |
| — | GPU Codename | GA104 (GN20-E6) |
| -Apple M3or above | Recommended CPU | -Intel Core i7 12700Hor above |
| 3478 | Steel Nomad Lite Score | 11488 |
| — | Time Spy | 11331 |
| 13529 | Solar Bay | 49158 |
| — | Port Royal | 6850 |
| — | Fire Strike | 28619 |
| 8229 | Wild Life Extreme | 23071 |
| — | Night Raid | 95266 |
| 500 MHz | Base Clock | 510-1035 MHz (configurable) |
| 1380 MHz | Boost Clock | 1035-1485 MHz (configurable) |
| 1280 | Shading Units | 5632 |
| 80 | Texture Mapping Units (TMUs) | 176 |
| 40 | Render Output Units (ROPs) | 80 |
| 160 | Compute Units (Pipelines) | 44 |
| — | Tensor Cores | 176 |
| — | Ray-tracing Cores | 44 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 4MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |