Quick/Comparisons
The Nvidia RTX Pro 6000 Blackwell dominates in gaming performance with a score of 99, far surpassing the Apple M4 Max GPU (40-core)'s 38.
The Nvidia RTX Pro 6000 Blackwell takes the top spot in content creation with a perfect score of 100, outperforming the Apple M4 Max GPU (40-core)'s 48.
Although the Nvidia RTX Pro 6000 Blackwell falls short of perfection, its 93 score in AI/ML is still significantly higher than the Apple M4 Max GPU (40-core)'s 46.
38 | Gaming | 99 |
48 | Workstation | 100 |
46 | AI/ML |
Compare with something else?
Compare Apple M4 Max with another →93 |
87 | Energy Efficiency | 49 |
50 | Quick Comparison Final Score | 98 |
| Custom | Graphics Processor | GB202-300-A1 |
| 5120 | Cores | 24064 |
| 320 / 160 | TMUs / ROPs | 752 / 190 |
| 5183.27 | Blender GPU | 16644.8 |
| — | Forza Horizon 5 | 230 |
| — | The Witcher 3 | 304 |
| — | Counter-Strike 2 | 259 |
| — | Far Cry 6 | 198 |
| — | Hogwarts Legacy | 192 |
| — | Call of Duty | 301 |
| — | Ghost of Tsushima | 172 |
| — | Cyberpunk 2077 | 221 |
| — | Tomb Raider | 303 |
| — | Average FPS | 242 |
| — | 1080p High | 282 |
| — | 1080p Ultra | 242 |
| — | 1440p Ultra | 197 |
| — | 4K Ultra | 118 |
| — | Margin of Error | High |
| 116754 | GB6 Compute Score | 419470 |
| 195 img/sec | Background Blur | 366.1 img/sec |
| 125.9 img/sec | Face Detection | 248 img/sec |
| 4.82 Gpixels/sec | Horizon Detection | 23.3 Gpixels/sec |
| 7.47 Gpixels/sec | Edge Detection | 35.2 Gpixels/sec |
| 5.8 Gpixels/sec | Gaussian Blur | 45 Gpixels/sec |
| 1.06 Gpixels/sec | Feature Matching | 2.47 Gpixels/sec |
| 417.4 Gpixels/sec | Stereo Matching | 3040 Gpixels/sec |
| 17079.9 FPS | Particle Physics | 42574.6 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 37305 |
| — | G2D Mark | 1375 |
| — | DirectX 11 | 338 FPS |
| — | DirectX 12 | 153 FPS |
| — | GPU Compute | 26895 Ops/s |
| 23632 | GB6 ML Single Precision | 50053 |
| 26019 | GB6 ML Half Precision | 74036 |
| 24606 | GB6 ML Quantized | 37636 |
| 10302 | Image Classification (SP) | 16903 |
| 23763 | Image Segmentation (HP) | 48518 |
| 29314 | Image Super Resolution (Q) | 54618 |
| 42016 | Face Detection (HP) | 103224 |
| 105972 | Pose Estimation (Q) | 338587 |
| 2963 | Text Classification (SP) | 3894 |
| 6594 | Machine Translation (HP) | 7291 |
| 8935 | Object Detection (SP) | 26498 |
| 43143 | Depth Estimation (Q) | 90853 |
| 221563 | Style Transfer (SP) | 635289 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 252 GPixel/s | Pixel Fill Rate | 497 GPixel/s |
| 505 GTexel/s | Texture Fill Rate | 1968 GTexel/s |
| 16.2 TFLOPS | FLOPS (FP32) | 126 TFLOPS |
| System Shared | Memory Type | GDDR7 |
| — | Memory Size | 96 GB |
| 8533 MHz | Memory Clock | 1750 MHz |
| — | Effective Memory Speed | 28000 Mbps |
| 512-bit | Bus | 512-bit |
| No | ECC | Yes |
| 546 GB/s | Memory Bandwidth | 1792 GB/s |
| Custom | Interface | PCIe 5.0 x16 |
| 50 W | TGP | 600 W |
| TSMC | Manufacturing | TSMC |
| 3 nm | Fabrication Process | 5 nm |
| — | Die Size | 750 mm² |
| — | Transistor Count | 92.2 billion |
| — | Transistor Density | 122.93 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 30, 2024 | Released | March 18, 2026 |
| Laptop | Case | Desktop |
| Professional | Purpose | Professional |
| Mid-range | Segment | High-end |
| Apple M GPU | Architecture | Blackwell 2.0 |
| Custom | GPU Codename | GB202-300-A1 |
| -Apple M4 Max (16-Core)or above | Recommended CPU | -AMD Ryzen 9 9955HX3Dor above |
| 14528 | Steel Nomad Lite Score | 50516 |
| — | Time Spy | 42437 |
| 61324 | Solar Bay | 258808 |
| — | Port Royal | 39952 |
| — | Fire Strike | 72183 |
| 36631 | Wild Life Extreme | 110070 |
| — | Night Raid | 216213 |
| 500 MHz | Base Clock | 1590 MHz |
| 1578 MHz | Boost Clock | 2617 MHz |
| 5120 | Shading Units | 24064 |
| 320 | Texture Mapping Units (TMUs) | 752 |
| 160 | Render Output Units (ROPs) | 190 |
| 640 | Compute Units (Pipelines) | 188 |
| — | Tensor Cores | 752 |
| — | Ray-tracing Cores | 188 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 128MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |