Quick/Comparisons
8 | Gaming | 46 |
21 | Workstation | 49 |
7 | AI/ML |
Compare with something else?
Compare Apple M1 GPU with another →39 |
89 | Energy Efficiency | 35 |
25 | Quick Comparison Final Score | 46 |
| — | Graphics Processor | GA104 |
| 1024 | Cores | 6144 |
| 64 / 32 | TMUs / ROPs | 192 / 96 |
| 259.48 | Blender GPU | 3746.83 |
| — | Forza Horizon 5 | 104 |
| — | The Witcher 3 | 143 |
| — | Counter-Strike 2 | 198 |
| — | Far Cry 6 | 125 |
| — | Hogwarts Legacy | 85 |
| — | Call of Duty | 120 |
| — | Ghost of Tsushima | 81 |
| — | Cyberpunk 2077 | 102 |
| — | Tomb Raider | 161 |
| — | Average FPS | 124 |
| — | 1080p High | 150 |
| — | 1080p Ultra | 124 |
| — | 1440p Ultra | 93 |
| — | 4K Ultra | 54 |
| — | Margin of Error | Low |
| 20842 | GB6 Compute Score | 136096 |
| 38.3 img/sec | Background Blur | 197 img/sec |
| 23 img/sec | Face Detection | 128.1 img/sec |
| 0.76 Gpixels/sec | Horizon Detection | 6.56 Gpixels/sec |
| 1.33 Gpixels/sec | Edge Detection | 8.81 Gpixels/sec |
| 0.87 Gpixels/sec | Gaussian Blur | 6.47 Gpixels/sec |
| 0.27 Gpixels/sec | Feature Matching | 1.26 Gpixels/sec |
| 69.9 Gpixels/sec | Stereo Matching | 609.3 Gpixels/sec |
| 2766.4 FPS | Particle Physics | 18205 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 23260 |
| — | G2D Mark | 1061 |
| — | DirectX 11 | 192 FPS |
| — | DirectX 12 | 91 FPS |
| — | GPU Compute | 11452 Ops/s |
| 3284 | GB6 ML Single Precision | 22516 |
| 5418 | GB6 ML Half Precision | 36377 |
| 4395 | GB6 ML Quantized | 14215 |
| 1400 | Image Classification (SP) | 8615 |
| 2280 | Image Segmentation (HP) | 24659 |
| 3402 | Image Super Resolution (Q) | 20631 |
| 5638 | Face Detection (HP) | 43834 |
| 12027 | Pose Estimation (Q) | 98347 |
| 2324 | Text Classification (SP) | 3162 |
| 4296 | Machine Translation (HP) | 4911 |
| 1530 | Object Detection (SP) | 11498 |
| 7675 | Depth Estimation (Q) | 34719 |
| 17293 | Style Transfer (SP) | 196833 |
| Core ML | Framework | ONNX |
| CPU | Backend | DirectML |
| 41 GPixel/s | Pixel Fill Rate | 170 GPixel/s |
| 82 GTexel/s | Texture Fill Rate | 340 GTexel/s |
| 2.6 TFLOPS | FLOPS (FP32) | 21.7 TFLOPS |
| System Shared | Memory Type | GDDR6X |
| — | Memory Size | 8 GB |
| 4266 MHz | Memory Clock | 1188 MHz |
| 8532 Mbps | Effective Memory Speed | 19000 Mbps |
| 128-bit | Bus | 256-bit |
| No | ECC | No |
| 68.2 GB/s | Memory Bandwidth | 608.3 GB/s |
| Custom | Interface | PCIe 4.0 x16 |
| 15 W | TGP | 290 W |
| Intel | Manufacturing | Samsung |
| 5 nm | Fabrication Process | 8 nm |
| — | Die Size | 392 mm² |
| 11.6 billion | Transistor Count | 17 billion |
| — | Transistor Density | 43.37 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 3 |
| — | CUDA | 8.6 |
| No | Ray Tracing | Yes |
| No | DLSS | DLSS 2 |
| — | DisplayPort | 1.4a |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| November 20, 2020 | Released | June 10, 2021 |
| Laptop | Case | Desktop |
| Professional | Purpose | Gaming |
| Mid-range | Segment | Mid-range |
| Apple M GPU | Architecture | Ampere |
| — | GPU Codename | GA104 |
| -Apple M1or above | Recommended CPU | -Intel Core i7 12700Kor above |
| 1979 | Steel Nomad Lite Score | 15402 |
| — | Time Spy | 14880 |
| 6311 | Solar Bay | 69900 |
| — | Port Royal | 8884 |
| — | Fire Strike | 37238 |
| 4957 | Wild Life Extreme | 31911 |
| — | Night Raid | 134012 |
| 450 MHz | Base Clock | 1575 MHz |
| 1278 MHz | Boost Clock | 1770 MHz |
| 1024 | Shading Units | 6144 |
| 64 | Texture Mapping Units (TMUs) | 192 |
| 32 | Render Output Units (ROPs) | 96 |
| 128 | Compute Units (Pipelines) | 48 |
| — | Tensor Cores | 192 |
| No | Ray-tracing Cores | 48 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 4MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |