Quick/Comparisons
50 | Gaming | 51 |
59 | Workstation | 56 |
41 | AI/ML |
Compare with something else?
Compare Apple M3 Ultra with another →33 |
61 | Energy Efficiency | 41 |
52 | Quick Comparison Final Score | 49 |
| Custom | Graphics Processor | GA103 |
| 10240 | Cores | 7424 |
| 640 / 320 | TMUs / ROPs | 232 / 96 |
| 7360.18 | Blender GPU | 2925.42 |
| — | 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 | — |
| 147719 | GB6 Compute Score | 144566 |
| 204.5 img/sec | Background Blur | 222.1 img/sec |
| 126 img/sec | Face Detection | 132.2 img/sec |
| 6.53 Gpixels/sec | Horizon Detection | 6.83 Gpixels/sec |
| 10.7 Gpixels/sec | Edge Detection | 9.64 Gpixels/sec |
| 9.5 Gpixels/sec | Gaussian Blur | 7.81 Gpixels/sec |
| 1.05 Gpixels/sec | Feature Matching | 1.27 Gpixels/sec |
| 560.3 Gpixels/sec | Stereo Matching | 615.9 Gpixels/sec |
| 25267.8 FPS | Particle Physics | 18134.8 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 17167 |
| — | G2D Mark | 713 |
| — | DirectX 11 | 132 FPS |
| — | DirectX 12 | 84 FPS |
| — | GPU Compute | 7423 Ops/s |
| 21356 | GB6 ML Single Precision | 15704 |
| 23530 | GB6 ML Half Precision | 26550 |
| 21706 | GB6 ML Quantized | 9885 |
| 8094 | Image Classification (SP) | 6982 |
| 24595 | Image Segmentation (HP) | 15050 |
| 25190 | Image Super Resolution (Q) | 14412 |
| 38337 | Face Detection (HP) | 27475 |
| 111021 | Pose Estimation (Q) | 66491 |
| 2786 | Text Classification (SP) | 2852 |
| 4339 | Machine Translation (HP) | 3306 |
| 7668 | Object Detection (SP) | 8121 |
| 37488 | Depth Estimation (Q) | 24442 |
| 244483 | Style Transfer (SP) | 156421 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 442 GPixel/s | Pixel Fill Rate | 101-160 GPixel/s |
| 883 GTexel/s | Texture Fill Rate | 244-386 GTexel/s |
| 28.3 TFLOPS | FLOPS (FP32) | 15.6-24.7 TFLOPS |
| System Shared | Memory Type | GDDR6 |
| — | Memory Size | 16 GB |
| 6400 MHz | Memory Clock | 2000 MHz |
| — | Effective Memory Speed | 16000 Mbps |
| 1024-bit | Bus | 256-bit |
| No | ECC | Yes |
| 819.3 GB/s | Memory Bandwidth | 512 GB/s |
| Custom | Interface | PCIe 4.0 x16 |
| 140 W | TGP | 80-165 W (configurable) |
| TSMC | Manufacturing | Samsung |
| 3 nm | Fabrication Process | 8 nm |
| — | Die Size | 496 mm² |
| 184 billion | Transistor Count | 22 billion |
| — | Transistor Density | 44.35 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 3 |
| — | CUDA | 8.6 |
| Yes | Ray Tracing | Yes |
| No | DLSS | DLSS 2 |
| — | DisplayPort | 1.4 |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| March 5, 2025 | Released | March 30, 2022 |
| Desktop | Case | Laptop |
| Professional | Purpose | Professional |
| High-end | Segment | High-end |
| Apple M GPU | Architecture | Ampere |
| Custom | GPU Codename | GA103 |
| -Apple M3 Ultraor above | Recommended CPU | -Intel Core i9 12900HXor above |
| 19491 | Steel Nomad Lite Score | — |
| — | Time Spy | 8880 |
| 73690 | Solar Bay | 32222 |
| — | Port Royal | 5730 |
| — | Fire Strike | 24599 |
| 44759 | Wild Life Extreme | 16879 |
| — | Night Raid | 27155 |
| — | Base Clock | — |
| 1380 MHz | Boost Clock | 1050-1663 MHz (configurable) |
| 10240 | Shading Units | 7424 |
| 640 | Texture Mapping Units (TMUs) | 232 |
| 320 | Render Output Units (ROPs) | 96 |
| 1280 | Compute Units (Pipelines) | 58 |
| — | Tensor Cores | 232 |
| — | Ray-tracing Cores | 58 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 4MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |