Quick/Comparisons
50 | Gaming | 32 |
59 | Workstation | 33 |
41 | AI/ML |
Compare with something else?
Compare Apple M3 Ultra with another →41 |
61 | Energy Efficiency | 66 |
52 | Quick Comparison Final Score | 39 |
| Custom | Graphics Processor | Strix Halo |
| 10240 | Cores | 2560 |
| 640 / 320 | TMUs / ROPs | 160 / 64 |
| 7360.18 | Blender GPU | 1335.18 |
| — | Forza Horizon 5 | 77 |
| — | The Witcher 3 | 101 |
| — | Counter-Strike 2 | 138 |
| — | Far Cry 6 | 92 |
| — | Hogwarts Legacy | 61 |
| — | Call of Duty | 89 |
| — | Ghost of Tsushima | 64 |
| — | Cyberpunk 2077 | 65 |
| — | Tomb Raider | 115 |
| — | Average FPS | 89 |
| — | 1080p High | 114 |
| — | 1080p Ultra | 89 |
| — | 1440p Ultra | 64 |
| — | 4K Ultra | 34 |
| — | Margin of Error | Low |
| 147719 | GB6 Compute Score | 93515 |
| 204.5 img/sec | Background Blur | 232.1 img/sec |
| 126 img/sec | Face Detection | 122.5 img/sec |
| 6.53 Gpixels/sec | Horizon Detection | 2.92 Gpixels/sec |
| 10.7 Gpixels/sec | Edge Detection | 4.07 Gpixels/sec |
| 9.5 Gpixels/sec | Gaussian Blur | 3.49 Gpixels/sec |
| 1.05 Gpixels/sec | Feature Matching | 1.08 Gpixels/sec |
| 560.3 Gpixels/sec | Stereo Matching | 396.3 Gpixels/sec |
| 25267.8 FPS | Particle Physics | 13016.5 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 18115 |
| — | G2D Mark | 1172 |
| — | DirectX 11 | 157 FPS |
| — | DirectX 12 | 84 FPS |
| — | GPU Compute | 9062 Ops/s |
| 21356 | GB6 ML Single Precision | 22565 |
| 23530 | GB6 ML Half Precision | 30098 |
| 21706 | GB6 ML Quantized | 18100 |
| 8094 | Image Classification (SP) | 11176 |
| 24595 | Image Segmentation (HP) | 14027 |
| 25190 | Image Super Resolution (Q) | 27198 |
| 38337 | Face Detection (HP) | 40613 |
| 111021 | Pose Estimation (Q) | 92878 |
| 2786 | Text Classification (SP) | 4158 |
| 4339 | Machine Translation (HP) | 6473 |
| 7668 | Object Detection (SP) | 7431 |
| 37488 | Depth Estimation (Q) | 41825 |
| 244483 | Style Transfer (SP) | 190221 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 442 GPixel/s | Pixel Fill Rate | 186 GPixel/s |
| 883 GTexel/s | Texture Fill Rate | 464 GTexel/s |
| 28.3 TFLOPS | FLOPS (FP32) | 14.8 TFLOPS |
| System Shared | Memory Type | System Shared |
| — | Memory Size | — |
| 6400 MHz | Memory Clock | — |
| — | Effective Memory Speed | — |
| 1024-bit | Bus | — |
| No | ECC | — |
| 819.3 GB/s | Memory Bandwidth | — |
| Custom | Interface | PCIe 5.0 x16 |
| 140 W | TGP | 45-120 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 4 nm |
| — | Die Size | 308 mm² |
| 184 billion | Transistor Count | — |
| — | Transistor Density | — |
| — | DirectX | 12.2 |
| — | Vulkan | 1.4 |
| — | OpenGL | 4.6 |
| — | OpenCL | 2.1 |
| — | CUDA | No |
| Yes | Ray Tracing | Yes |
| No | DLSS | No |
| — | DisplayPort | 2.1a |
| Apple | Vendor | Amd |
| Integrated | Build | Integrated |
| March 5, 2025 | Released | January 7, 2025 |
| Desktop | Case | Laptop |
| Professional | Purpose | Generic |
| High-end | Segment | Mid-range |
| Apple M GPU | Architecture | RDNA 3.5 |
| Custom | GPU Codename | Strix Halo |
| -Apple M3 Ultraor above | Recommended CPU | -AMD Ryzen AI Max+ 395or above |
| 19491 | Steel Nomad Lite Score | 9818 |
| — | Time Spy | 10010 |
| 73690 | Solar Bay | 46934 |
| — | Port Royal | 5512 |
| — | Fire Strike | 26722 |
| 44759 | Wild Life Extreme | 19043 |
| — | Night Raid | 100428 |
| — | Base Clock | 1295 MHz |
| 1380 MHz | Boost Clock | 2900 MHz |
| 10240 | Shading Units | 2560 |
| 640 | Texture Mapping Units (TMUs) | 160 |
| 320 | Render Output Units (ROPs) | 64 |
| 1280 | Compute Units (Pipelines) | 40 |
| — | Tensor Cores | No |
| — | Ray-tracing Cores | 40 |
| — | L1 Cache | 512KB per cluster |
| — | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |