Quick/Comparisons
The Nvidia RTX 5000 Ada outperforms the Apple M2 Max GPU (38-core) by a significant margin of 47 points, making it the clear winner for gaming. Its strong performance is due to its high score of 73.
The Nvidia RTX 5000 Ada takes the lead in content creation with a substantial 43-point gap, thanks to its strong workstation performance and large 32GB memory pool. Its score of 82 far surpasses the Apple M2 Max GPU (38-core)'s score of 39.
The Nvidia RTX 5000 Ada edges out the Apple M2 Max GPU (38-core) in AI/ML tasks with a 25-point difference, driven by its Ada Lovelace architecture and 32GB GDDR6 memory. Although the gap is not as large as in other use cases, the Nvidia RTX 5000 Ada still emerges as the winner.
26 | Gaming | 73 |
39 | Workstation | 82 |
29 | AI/ML | 54 |
76 | Energy Efficiency | 45 |
38 | Quick Comparison Final Score | 73 |
| — | Graphics Processor | AD102 |
| 4864 | Cores | 12800 |
| 304 / 152 | TMUs / ROPs | 400 / 176 |
| 1777.05 | Blender GPU | 7348.98 |
| — | Forza Horizon 5 | 146 |
| — | The Witcher 3 | 186 |
| — | Counter-Strike 2 | 213 |
| — | Far Cry 6 | 151 |
| — | Hogwarts Legacy | 117 |
| — | Call of Duty | 179 |
| — | Ghost of Tsushima | 108 |
| — | Cyberpunk 2077 | 148 |
| — | Tomb Raider | 242 |
| — | Average FPS | 166 |
| — | 1080p High | 196 |
| — | 1080p Ultra | 166 |
| — | 1440p Ultra | 132 |
| — | 4K Ultra | 78 |
| — | Margin of Error | High |
| 86936 | GB6 Compute Score | 260099 |
| 149 img/sec | Background Blur | 286.6 img/sec |
| 98.1 img/sec | Face Detection | 220.4 img/sec |
| 3.8 Gpixels/sec | Horizon Detection | 9.93 Gpixels/sec |
| 6.88 Gpixels/sec | Edge Detection | 13.7 Gpixels/sec |
| 4.44 Gpixels/sec | Gaussian Blur | 18.9 Gpixels/sec |
| 0.73 Gpixels/sec | Feature Matching | 2.2 Gpixels/sec |
| 264.9 Gpixels/sec | Stereo Matching | 1580 Gpixels/sec |
| 11144.2 FPS | Particle Physics | 41235.6 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 30196 |
| — | G2D Mark | 998 |
| — | DirectX 11 | 273 FPS |
| — | DirectX 12 | 106 FPS |
| — | GPU Compute | 20736 Ops/s |
| 15005 | GB6 ML Single Precision | 24715 |
| 16456 | GB6 ML Half Precision | 36753 |
| 15418 | GB6 ML Quantized | 19649 |
| 6657 | Image Classification (SP) | 9943 |
| 13854 | Image Segmentation (HP) | 19696 |
| 18642 | Image Super Resolution (Q) | 32252 |
| 27965 | Face Detection (HP) | 43098 |
| 83144 | Pose Estimation (Q) | 138689 |
| 2377 | Text Classification (SP) | 2893 |
| 2052 | Machine Translation (HP) | 4064 |
| 6495 | Object Detection (SP) | 12467 |
| 31105 | Depth Estimation (Q) | 50352 |
| 135686 | Style Transfer (SP) | 265482 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 212 GPixel/s | Pixel Fill Rate | 449 GPixel/s |
| 425 GTexel/s | Texture Fill Rate | 1020 GTexel/s |
| 13.6 TFLOPS | FLOPS (FP32) | 65.3 TFLOPS |
| System Shared | Memory Type | GDDR6 |
| — | Memory Size | 32 GB |
| 6400 MHz | Memory Clock | 2250 MHz |
| 12800 Mbps | Effective Memory Speed | 18000 Mbps |
| 512-bit | Bus | 256-bit |
| No | ECC | No |
| 409.6 GB/s | Memory Bandwidth | 576 GB/s |
| Custom | Interface | PCIe 4.0 x16 |
| 70 W | TGP | 250 W |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 5 nm |
| — | Die Size | 609 mm² |
| 52 billion | Transistor Count | 76 billion |
| — | Transistor Density | 124.79 MTr/mm² |
| — | DirectX | 12 |
| — | Vulkan | 1.3 |
| — | OpenGL | 4.6 |
| — | OpenCL | 3 |
| — | CUDA | 8.9 |
| No | Ray Tracing | Yes |
| No | DLSS | DLSS 3 |
| — | DisplayPort | 1.4a |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| January 17, 2023 | Released | August 9, 2023 |
| Laptop | Case | Desktop |
| Professional | Purpose | Professional |
| High-end | Segment | High-end |
| Apple M GPU | Architecture | Ada Lovelace |
| — | GPU Codename | AD102 |
| — | Recommended CPU | -Intel Core i9 14900Kor above |
| 9413 | Steel Nomad Lite Score | 29190 |
| — | Time Spy | 18675 |
| 30427 | Solar Bay | 128428 |
| — | Port Royal | 17754 |
| — | Fire Strike | 62027 |
| 24969 | Wild Life Extreme | 52866 |
| — | Night Raid | 210443 |
| 450 MHz | Base Clock | 1155 MHz |
| 1398 MHz | Boost Clock | 2550 MHz |
| 4864 | Shading Units | 12800 |
| 304 | Texture Mapping Units (TMUs) | 400 |
| 152 | Render Output Units (ROPs) | 176 |
| 608 | Compute Units (Pipelines) | 100 |
| — | Tensor Cores | 400 |
| — | Ray-tracing Cores | 100 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 72MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare Apple M2 Max with another →