Quick/Comparisons
14 | Gaming | 57 |
23 | Workstation | 55 |
10 | AI/ML |
Compare with something else?
Compare Apple M3 GPU with another →48 |
95 | Energy Efficiency | 45 |
30 | Quick Comparison Final Score | 53 |
| — | Graphics Processor | AD104 (GN21-X9) |
| 1280 | Cores | 7424 |
| 80 / 40 | TMUs / ROPs | 232 / 80 |
| 155.27 | Blender GPU | 5463.87 |
| — | Forza Horizon 5 | 120 |
| — | The Witcher 3 | 163 |
| — | Counter-Strike 2 | 209 |
| — | Far Cry 6 | 139 |
| — | Hogwarts Legacy | 98 |
| — | Call of Duty | 153 |
| — | Ghost of Tsushima | 92 |
| — | Cyberpunk 2077 | 122 |
| — | Tomb Raider | 208 |
| — | Average FPS | 145 |
| — | 1080p High | 171 |
| — | 1080p Ultra | 145 |
| — | 1440p Ultra | 112 |
| — | 4K Ultra | 66 |
| — | Margin of Error | Low |
| 30027 | GB6 Compute Score | 170767 |
| 60.3 img/sec | Background Blur | 328.9 img/sec |
| 40.8 img/sec | Face Detection | 192.3 img/sec |
| 1.21 Gpixels/sec | Horizon Detection | 6.51 Gpixels/sec |
| 1.64 Gpixels/sec | Edge Detection | 9.11 Gpixels/sec |
| 1.2 Gpixels/sec | Gaussian Blur | 10.1 Gpixels/sec |
| 0.37 Gpixels/sec | Feature Matching | 1.56 Gpixels/sec |
| 102.6 Gpixels/sec | Stereo Matching | 660.7 Gpixels/sec |
| 3293.3 FPS | Particle Physics | 20765.5 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 24711 |
| — | G2D Mark | 922 |
| — | DirectX 11 | 241 FPS |
| — | DirectX 12 | 96 FPS |
| — | GPU Compute | 11008 Ops/s |
| 4120 | GB6 ML Single Precision | 26637 |
| 6992 | GB6 ML Half Precision | 39211 |
| 5638 | GB6 ML Quantized | 19995 |
| 1579 | Image Classification (SP) | 10385 |
| 2944 | Image Segmentation (HP) | 25896 |
| 4425 | Image Super Resolution (Q) | 27816 |
| 6661 | Face Detection (HP) | 51976 |
| 15020 | Pose Estimation (Q) | 142459 |
| 2984 | Text Classification (SP) | 3405 |
| 5876 | Machine Translation (HP) | 4860 |
| 1826 | Object Detection (SP) | 13308 |
| 9339 | Depth Estimation (Q) | 49203 |
| 22291 | Style Transfer (SP) | 262792 |
| Core ML | Framework | ONNX |
| CPU | Backend | DirectML |
| 55 GPixel/s | Pixel Fill Rate | 133-182 GPixel/s |
| 110 GTexel/s | Texture Fill Rate | 386-529 GTexel/s |
| 3.5 TFLOPS | FLOPS (FP32) | 24.7-33.9 TFLOPS |
| System Shared | Memory Type | GDDR6 |
| — | Memory Size | 12 GB |
| 6400 MHz | Memory Clock | 2250 MHz |
| 12800 Mbps | Effective Memory Speed | 18000 Mbps |
| 128-bit | Bus | 192-bit |
| No | ECC | No |
| 102.4 GB/s | Memory Bandwidth | 432 GB/s |
| Custom | Interface | PCIe 4.0 x16 |
| 15 W | TGP | 60-175 W (configurable) |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 5 nm |
| — | Die Size | 295 mm² |
| 13.8 billion | Transistor Count | 35.8 billion |
| — | Transistor Density | 121.36 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 3 |
| — | CUDA | 8.9 |
| Yes | Ray Tracing | Yes |
| No | DLSS | DLSS 3 |
| — | DisplayPort | 1.4a |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| October 31, 2023 | Released | February 8, 2023 |
| Laptop | Case | Laptop |
| Professional | Purpose | Gaming |
| Mid-range | Segment | High-end |
| Apple M GPU | Architecture | Ada Lovelace |
| — | GPU Codename | AD104 (GN21-X9) |
| -Apple M3or above | Recommended CPU | -Intel Core Ultra 9 185Hor above |
| 3478 | Steel Nomad Lite Score | 19263 |
| — | Time Spy | 18832 |
| 13529 | Solar Bay | 85404 |
| — | Port Royal | 11837 |
| — | Fire Strike | 41182 |
| 8229 | Wild Life Extreme | 36965 |
| — | Night Raid | 147212 |
| 500 MHz | Base Clock | 1290-1860 MHz (configurable) |
| 1380 MHz | Boost Clock | 1665-2280 MHz (configurable) |
| 1280 | Shading Units | 7424 |
| 80 | Texture Mapping Units (TMUs) | 232 |
| 40 | Render Output Units (ROPs) | 80 |
| 160 | Compute Units (Pipelines) | 58 |
| — | Tensor Cores | 232 |
| — | Ray-tracing Cores | 58 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 48MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |