Quick/Comparisons
19 | Gaming | 20 |
27 | Workstation | 27 |
24 | AI/ML |
Compare with something else?
Compare Apple M3 Pro with another →15 |
90 | Energy Efficiency | 28 |
34 | Quick Comparison Final Score | 22 |
| Custom | Graphics Processor | GP104 |
| 2304 | Cores | 1920 |
| 144 / 72 | TMUs / ROPs | 120 / 64 |
| 1678.22 | Blender GPU | 557.8 |
| — | Forza Horizon 5 | 53 |
| — | The Witcher 3 | 69 |
| — | Counter-Strike 2 | 96 |
| — | Far Cry 6 | 65 |
| — | Hogwarts Legacy | 39 |
| — | Call of Duty | 49 |
| — | Ghost of Tsushima | 35 |
| — | Cyberpunk 2077 | 40 |
| — | Tomb Raider | 72 |
| — | Average FPS | 58 |
| — | 1080p High | 71 |
| — | 1080p Ultra | 58 |
| — | 1440p Ultra | 41 |
| — | 4K Ultra | 20 |
| — | Margin of Error | Low |
| 43099 | GB6 Compute Score | 53429 |
| 82.1 img/sec | Background Blur | 142.3 img/sec |
| 53.6 img/sec | Face Detection | 65.1 img/sec |
| 1.82 Gpixels/sec | Horizon Detection | 2.2 Gpixels/sec |
| 2.59 Gpixels/sec | Edge Detection | 3.23 Gpixels/sec |
| 1.84 Gpixels/sec | Gaussian Blur | 3.74 Gpixels/sec |
| 0.48 Gpixels/sec | Feature Matching | 0.3 Gpixels/sec |
| 148.6 Gpixels/sec | Stereo Matching | 195.6 Gpixels/sec |
| 4893.9 FPS | Particle Physics | 5161.6 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 13508 |
| — | G2D Mark | 847 |
| — | DirectX 11 | 100 FPS |
| — | DirectX 12 | 48 FPS |
| — | GPU Compute | 6118 Ops/s |
| 11804 | GB6 ML Single Precision | 8330 |
| 13606 | GB6 ML Half Precision | 8293 |
| 12673 | GB6 ML Quantized | 6176 |
| 5578 | Image Classification (SP) | 2295 |
| 9924 | Image Segmentation (HP) | 4955 |
| 13323 | Image Super Resolution (Q) | 10587 |
| 19900 | Face Detection (HP) | 8747 |
| 35730 | Pose Estimation (Q) | 51263 |
| 2904 | Text Classification (SP) | 1461 |
| 4985 | Machine Translation (HP) | 2342 |
| 5496 | Object Detection (SP) | 2960 |
| 23344 | Depth Estimation (Q) | 15275 |
| 78342 | Style Transfer (SP) | 104907 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 99 GPixel/s | Pixel Fill Rate | 108 GPixel/s |
| 199 GTexel/s | Texture Fill Rate | 202 GTexel/s |
| 6.4 TFLOPS | FLOPS (FP32) | 6.5 TFLOPS |
| System Shared | Memory Type | GDDR5 |
| — | Memory Size | 8 GB |
| 6400 MHz | Memory Clock | 2002 MHz |
| 12800 Mbps | Effective Memory Speed | 8000 Mbps |
| 192-bit | Bus | 256-bit |
| No | ECC | No |
| 153.6 GB/s | Memory Bandwidth | 256.3 GB/s |
| Custom | Interface | PCIe 3.0 x16 |
| 30 W | TGP | 150 W |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 16 nm |
| — | Die Size | 314 mm² |
| 25.2 billion | Transistor Count | 7 billion |
| — | Transistor Density | 22.29 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 3 |
| — | CUDA | 6.1 |
| Yes | Ray Tracing | No |
| No | DLSS | No |
| — | DisplayPort | 1.4a |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| October 31, 2023 | Released | January 10, 2016 |
| Laptop | Case | Desktop |
| Professional | Purpose | Gaming |
| Mid-range | Segment | Mid-range |
| Apple M GPU | Architecture | Pascal |
| Custom | GPU Codename | GP104 |
| -Apple M3 Proor above | Recommended CPU | -Intel Core i7 7700Kor above |
| 6092 | Steel Nomad Lite Score | 5857 |
| — | Time Spy | 6091 |
| 22577 | Solar Bay | 16018 |
| — | Port Royal | 1282 |
| — | Fire Strike | 18706 |
| 14001 | Wild Life Extreme | 12190 |
| — | Night Raid | 72438 |
| 500 MHz | Base Clock | 1506 MHz |
| 1380 MHz | Boost Clock | 1683 MHz |
| 2304 | Shading Units | 1920 |
| 144 | Texture Mapping Units (TMUs) | 120 |
| 72 | Render Output Units (ROPs) | 64 |
| 288 | Compute Units (Pipelines) | 15 |
| — | Tensor Cores | No |
| — | Ray-tracing Cores | No |
| — | L1 Cache | 48KB per cluster |
| — | L2 Cache | 2MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |