Quick/Comparisons
The GeForce RTX 5070 Ti's superior gaming performance, with a 45-point lead, makes it the clear winner in this use case. Its 76 score is significantly higher than the Apple M1 Ultra GPU's 31.
Both GPUs have strong workstation capabilities, with the GeForce RTX 5070 Ti edging out the Apple M1 Ultra GPU by just 24 points. However, the score gap is within the 5-point threshold, making this a tie.
The GeForce RTX 5070 Ti's 41-point lead in AI/ML capabilities makes it the clear winner in this use case. Its 66 score is significantly higher than the Apple M1 Ultra GPU's 25.
31 | Gaming | 76 |
50 | Workstation | 74 |
25 | AI/ML | 66 |
59 | Energy Efficiency | 49 |
40 | Quick Comparison Final Score | 74 |
| Custom | Graphics Processor | GB203-200-A1 |
| 8192 | Cores | 8960 |
| 512 / 256 | TMUs / ROPs | 280 / 96 |
| 1644.94 | Blender GPU | 7584.67 |
| — | Cyberpunk 2077 | 174 |
| — | Forza Horizon 5 | 166 |
| — | The Witcher 3 | 213 |
| — | Counter-Strike 2 | 219 |
| — | Far Cry 6 | 159 |
| — | Hogwarts Legacy | 135 |
| — | Call of Duty | 205 |
| — | Ghost of Tsushima | 121 |
| — | Tomb Raider | 260 |
| — | Average FPS | 184 |
| — | 1080p High | 217 |
| — | 1080p Ultra | 184 |
| — | 1440p Ultra | 148 |
| — | 4K Ultra | 86 |
| — | Margin of Error | Low |
| 106397 | GB6 Compute Score | 240508 |
| 157.3 img/sec | Background Blur | 297.9 img/sec |
| 108.7 img/sec | Face Detection | 218.6 img/sec |
| 5.07 Gpixels/sec | Horizon Detection | 10.1 Gpixels/sec |
| 10.7 Gpixels/sec | Edge Detection | 13.9 Gpixels/sec |
| 6.59 Gpixels/sec | Gaussian Blur | 14.8 Gpixels/sec |
| 0.73 Gpixels/sec | Feature Matching | 2.03 Gpixels/sec |
| 293.4 Gpixels/sec | Stereo Matching | 1290 Gpixels/sec |
| 14154.4 FPS | Particle Physics | 35438.3 FPS |
| OpenCL | API | OpenCL |
| — | G3D Mark Score | 32425 |
| — | G2D Mark | 1329 |
| — | DirectX 11 | 304 FPS |
| — | DirectX 12 | 126 FPS |
| — | GPU Compute | 18089 Ops/s |
| 13120 | GB6 ML Single Precision | 36504 |
| 14103 | GB6 ML Half Precision | 64459 |
| 13229 | GB6 ML Quantized | 27312 |
| 4834 | Image Classification (SP) | 14913 |
| 14458 | Image Segmentation (HP) | 45199 |
| 13757 | Image Super Resolution (Q) | 39360 |
| 23618 | Face Detection (HP) | 84188 |
| 80481 | Pose Estimation (Q) | 161400 |
| 2265 | Text Classification (SP) | 4010 |
| 1798 | Machine Translation (HP) | 7973 |
| 4975 | Object Detection (SP) | 18934 |
| 25755 | Depth Estimation (Q) | 67761 |
| 135609 | Style Transfer (SP) | 335879 |
| Core ML | Framework | ONNX |
| GPU | Backend | DirectML |
| 332 GPixel/s | Pixel Fill Rate | 235 GPixel/s |
| 664 GTexel/s | Texture Fill Rate | 687 GTexel/s |
| 21.2 TFLOPS | FLOPS (FP32) | 43.9 TFLOPS |
| System Shared | Memory Type | GDDR7 |
| — | Memory Size | 16 GB |
| 6400 MHz | Memory Clock | 1750 MHz |
| 12800 Mbps | Effective Memory Speed | 28000 Mbps |
| 1024-bit | Bus | 256-bit |
| No | ECC | No |
| 819.2 GB/s | Memory Bandwidth | 896 GB/s |
| Custom | Interface | PCIe 5.0 x16 |
| 120 W | TGP | 300 W |
| TSMC | Manufacturing | TSMC |
| 5 nm | Fabrication Process | 4 nm |
| — | Die Size | 378 mm² |
| 92.8 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 |
| No | Ray Tracing | Yes |
| No | DLSS | DLSS 4 |
| — | DisplayPort | 2.1b |
| Apple | Vendor | Nvidia |
| Integrated | Build | Discrete |
| March 8, 2022 | Released | January 7, 2025 |
| Laptop | Case | Desktop |
| Professional | Purpose | Gaming |
| High-end | Segment | Mid-range |
| Apple M GPU | Architecture | Blackwell 2.0 |
| Custom | GPU Codename | GB203-200-A1 |
| -Apple M1 Ultraor above | Recommended CPU | -Intel Core Ultra 5 245Kor above |
| 10731 | Steel Nomad Lite Score | 29989 |
| — | Time Spy | 27815 |
| 36712 | Solar Bay | 136964 |
| — | Port Royal | 19613 |
| — | Fire Strike | 67382 |
| 29343 | Wild Life Extreme | 57734 |
| — | Night Raid | 189734 |
| 450 MHz | Base Clock | 2295 MHz |
| 1296 MHz | Boost Clock | 2452 MHz |
| 8192 | Shading Units | 8960 |
| 512 | Texture Mapping Units (TMUs) | 280 |
| 256 | Render Output Units (ROPs) | 96 |
| 1024 | Compute Units (Pipelines) | 70 |
| — | Tensor Cores | 280 |
| — | Ray-tracing Cores | 70 |
| — | L1 Cache | 128KB per cluster |
| — | L2 Cache | 48MB shared |
| 2 IPC | Instructions Per Cycle | 2 IPC |
Compare with something else?
Compare Apple M1 Ultra with another →