Quick/Comparisons
33 | Gaming | 55 |
42 | Workstation | 56 |
39 | AI/ML |
Compare with something else?
Compare Apple M3 Max with another →43 |
82 | Energy Efficiency | 35 |
44 | Quick Comparison Final Score | 54 |
| — | Graphics Processor | GA102 |
| 5120 | Cores | 8704 |
| 320 / 160 | TMUs / ROPs | 272 / 96 |
| 3797.82 | Blender GPU | 4812.92 |
| — | Forza Horizon 5 | 114 |
| — | The Witcher 3 | 163 |
| — | Counter-Strike 2 | 205 |
| — | Far Cry 6 | 137 |
| — | Hogwarts Legacy | 97 |
| — | Call of Duty | 146 |
| — | Ghost of Tsushima | 90 |
| — | Cyberpunk 2077 | 121 |
| — | Tomb Raider | 193 |
| — | Average FPS | 141 |
| — | 1080p High | 167 |
| — | 1080p Ultra | 141 |
| — | 1440p Ultra | 109 |
| — | 4K Ultra | 63 |
| — | Margin of Error | Low |
| 94528 | GB6 Compute Score | 161903 |
| 157.6 img/sec | Background Blur | 217.3 img/sec |
| 100.9 img/sec | Face Detection | 141.9 img/sec |
| 3.99 Gpixels/sec | Horizon Detection | 8.19 Gpixels/sec |
| 6.04 Gpixels/sec | Edge Detection | 11 Gpixels/sec |
| 5 Gpixels/sec | Gaussian Blur | 8.43 Gpixels/sec |
| 0.84 Gpixels/sec | Feature Matching | 1.29 Gpixels/sec |
| 363.8 Gpixels/sec | Stereo Matching | 780.7 Gpixels/sec |
| 12226.9 FPS | Particle Physics | 22538.5 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 25030 |
| — | G2D Mark | 1062 |
| — | DirectX 11 | 207 FPS |
| — | DirectX 12 | 100 FPS |
| — | GPU Compute | 14154 Ops/s |
| 20068 | GB6 ML Single Precision | 24003 |
| 22390 | GB6 ML Half Precision | 36895 |
| 20681 | GB6 ML Quantized | 14629 |
| 8685 | Image Classification (SP) | 9439 |
| 20739 | Image Segmentation (HP) | 25471 |
| 22775 | Image Super Resolution (Q) | 19892 |
| 37676 | Face Detection (HP) | 46956 |
| 91609 | Pose Estimation (Q) | 104704 |
| 2909 | Text Classification (SP) | 2712 |
| 5274 | Machine Translation (HP) | 4828 |
| 7480 | Object Detection (SP) | 12153 |
| 38212 | Depth Estimation (Q) | 36822 |
| 190681 | Style Transfer (SP) | 246005 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 221 GPixel/s | Pixel Fill Rate | 164 GPixel/s |
| 442 GTexel/s | Texture Fill Rate | 465 GTexel/s |
| 14.1 TFLOPS | FLOPS (FP32) | 29.8 TFLOPS |
| System Shared | Memory Type | GDDR6X |
| — | Memory Size | 10 GB |
| 6400 MHz | Memory Clock | 1188 MHz |
| 12800 Mbps | Effective Memory Speed | 19000 Mbps |
| 512-bit | Bus | 320-bit |
| No | ECC | No |
| 409.6 GB/s | Memory Bandwidth | 760.3 GB/s |
| Custom | Interface | PCIe 4.0 x16 |
| 60 W | TGP | 320 W |
| TSMC | Manufacturing | Samsung |
| 3 nm | Fabrication Process | 8 nm |
| — | Die Size | 628 mm² |
| 56 billion | Transistor Count | 28 billion |
| — | Transistor Density | 44.59 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | 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 | September 1, 2020 |
| Laptop | Case | Desktop |
| Professional | Purpose | Gaming |
| High-end | Segment | High-end |
| Apple M GPU | Architecture | Ampere |
| — | GPU Codename | GA102 |
| -Apple M3 Maxor above | Recommended CPU | -Intel Core i7 12700Kor above |
| 12391 | Steel Nomad Lite Score | 19046 |
| — | Time Spy | 17664 |
| 49656 | Solar Bay | 85890 |
| — | Port Royal | 11597 |
| — | Fire Strike | 42468 |
| 30787 | Wild Life Extreme | 37899 |
| — | Night Raid | 132892 |
| 500 MHz | Base Clock | 1440 MHz |
| 1380 MHz | Boost Clock | 1710 MHz |
| 5120 | Shading Units | 8704 |
| 320 | Texture Mapping Units (TMUs) | 272 |
| 160 | Render Output Units (ROPs) | 96 |
| 640 | Compute Units (Pipelines) | 68 |
| — | Tensor Cores | 272 |
| — | Ray-tracing Cores | 68 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 5MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |