Quick/Comparisons
19 | Gaming | 56 |
32 | Workstation | 67 |
34 | AI/ML |
Compare with something else?
Compare Apple M5 GPU with another →53 |
100 | Energy Efficiency | 49 |
38 | Quick Comparison Final Score | 57 |
| Custom | Graphics Processor | GB203-300-A1 |
| 1280 | Cores | 7680 |
| 80 / 40 | TMUs / ROPs | 240 / 96 |
| 264.45 | Blender GPU | 5699.95 |
| — | Forza Horizon 5 | 114 |
| — | The Witcher 3 | 157 |
| — | Counter-Strike 2 | 204 |
| — | Far Cry 6 | 132 |
| — | Hogwarts Legacy | 96 |
| — | Call of Duty | 143 |
| — | Ghost of Tsushima | 87 |
| — | Cyberpunk 2077 | 118 |
| — | Tomb Raider | 197 |
| — | Average FPS | 139 |
| — | 1080p High | 166 |
| — | 1080p Ultra | 139 |
| — | 1440p Ultra | 106 |
| — | 4K Ultra | 63 |
| — | Margin of Error | High |
| 75777 | GB6 Compute Score | 197909 |
| — | Background Blur | 267.1 img/sec |
| — | Face Detection | 165.5 img/sec |
| — | Horizon Detection | 8.72 Gpixels/sec |
| — | Edge Detection | 12.3 Gpixels/sec |
| — | Gaussian Blur | 12.6 Gpixels/sec |
| — | Feature Matching | 1.83 Gpixels/sec |
| — | Stereo Matching | 942.6 Gpixels/sec |
| — | Particle Physics | 25637.1 FPS |
| — | API | OpenCL |
| — | G3D Mark Score | 23401 |
| — | G2D Mark | 874 |
| — | DirectX 11 | 232 FPS |
| — | DirectX 12 | 109 FPS |
| — | GPU Compute | 8154 Ops/s |
| 13289 | GB6 ML Single Precision | 27844 |
| 24800 | GB6 ML Half Precision | 48544 |
| 23572 | GB6 ML Quantized | 21478 |
| 6348 | Image Classification (SP) | 11723 |
| 16813 | Image Segmentation (HP) | 32142 |
| 30322 | Image Super Resolution (Q) | 30735 |
| 34387 | Face Detection (HP) | 58722 |
| 111591 | Pose Estimation (Q) | 129551 |
| 2319 | Text Classification (SP) | 3539 |
| 5555 | Machine Translation (HP) | 6242 |
| 6450 | Object Detection (SP) | 14495 |
| 53195 | Depth Estimation (Q) | 50690 |
| 82314 | Style Transfer (SP) | 231154 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 65 GPixel/s | Pixel Fill Rate | 163-180 GPixel/s |
| 130 GTexel/s | Texture Fill Rate | 408-450 GTexel/s |
| 4.1 TFLOPS | FLOPS (FP32) | 26.1-28.8 TFLOPS |
| System Shared | Memory Type | GDDR7 |
| — | Memory Size | 16 GB |
| 9600 MHz | Memory Clock | 1750 MHz |
| — | Effective Memory Speed | 28000 Mbps |
| 128-bit | Bus | 256-bit |
| No | ECC | No |
| 153 GB/s | Memory Bandwidth | 896 GB/s |
| Custom | Interface | PCIe 5.0 x16 |
| — | TGP | 80-175 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 5 nm |
| — | Die Size | 378 mm² |
| 28 billion | Transistor Count | 45.6 billion |
| — | Transistor Density | 120.63 MTr/mm² |
| — | DirectX | 12.2 |
| — | Vulkan | 1.4 |
| — | OpenGL | 4.6 |
| — | OpenCL | 3 |
| — | CUDA | 12 |
| Yes | Ray Tracing | Yes |
| No | DLSS | DLSS 4 |
| — | DisplayPort | 2.1b |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| October 15, 2025 | Released | March 18, 2025 |
| Laptop | Case | Laptop |
| Professional | Purpose | Professional |
| Mid-range | Segment | High-end |
| Apple M GPU | Architecture | Blackwell 2.0 |
| Custom | GPU Codename | GB203-300-A1 |
| -Apple M5 (10-Core)or above | Recommended CPU | -Intel Core Ultra 9 275HXor above |
| 5337 | Steel Nomad Lite Score | 22035 |
| — | Time Spy | 13961 |
| 24534 | Solar Bay | — |
| — | Port Royal | 12094 |
| — | Fire Strike | 38402 |
| 12266 | Wild Life Extreme | — |
| — | Night Raid | — |
| — | Base Clock | 975-1702 MHz (configurable) |
| 1620 MHz | Boost Clock | 1702-1875 MHz (configurable) |
| 1280 | Shading Units | 7680 |
| 80 | Texture Mapping Units (TMUs) | 240 |
| 40 | Render Output Units (ROPs) | 96 |
| 160 | Compute Units (Pipelines) | 60 |
| — | Tensor Cores | 240 |
| — | Ray-tracing Cores | 60 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 64MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |