Apple M3 Ultra GPU (80-core) vs Nvidia GeForce RTX 3090 Ti
Compared 2 processors: Apple M3 Ultra GPU (80-core), vs Nvidia GeForce RTX 3090 Ti. Below are specs, tests, pros and cons for each device
Review
Games
Performance in DirectX, OpenCL, Vulkan
AI/ML
AI workload testing
Professional Apps
3D modeling, video editing, rendering
Energy Efficiency
Power usage in various scenarios
Tests & specs
Specs & test comparisonGames
| Game title | 1080p (High) 1080p (Ultra) 1440p (Ultra) 4K (Ultra) | |
Forza Horizon 5 | 0 0 0 0 | 196 139 127 96 |
The Witcher 3 | 37 25 0 0 | 218 185 131 70 |
Counter-Strike 2 | 0 0 0 0 | 290 218 161 86 |
Far Cry 6 | 0 0 0 0 | 158 148 129 81 |
Hogwarts Legacy | 0 0 0 0 | 136 113 91 50 |
COD Modern Warfare 3 | 0 0 0 0 | 181 173 136 89 |
Ghost of Tsushima | 0 0 0 0 | 124 106 91 55 |
Cyberpunk 2077 | 0 0 0 0 | 163 146 98 45 |
Shadow of the Tomb Raider | 0 0 0 0 | 261 238 186 101 |
| 1080p High | n/a | 192 |
| 1080p Ultra | n/a | 163 |
| 1440p Ultra | n/a | 128 |
| 4K Ultra | n/a | 75 |
| Deviation | n/a | Low |
Benchmarks
| Time Spy | n/a | 21841 |
| Solar Bay | n/a | 109151 |
| Port Royal | n/a | 14882 |
| Fire Strike | n/a | 51985 |
| Night Raid | n/a | 155722 |
| Background Blur | 204.5 frames/s | 362.7 frames/s |
| Face Detection | 126 frames/s | 227.9 frames/s |
| Horizon Detection | 6.53 Gpix/sec | 11.3 Gpix/sec |
| Edge Detection | 10.7 Gpix/sec | 15.9 Gpix/sec |
| Gaussian Blur | 9.5 Gpix/sec | 12.1 Gpix/sec |
| Feature Matching | 1.05 Gpix/sec | 1.56 Gpix/sec |
| Stereo Matching | 560.3 Gpix/sec | 1080 Gpix/sec |
| Particle Physics | 25267.8 fps | 31024.6 fps |
| API | OpenCL | OpenCL |
| G2D Mark | n/a | 1227 |
| DirectX 11 | n/a | 241 fps |
| DirectX 12 | n/a | 122 fps |
| GPU Compute | n/a | 17022 Ops/s |
| Aztec Ruins (Normal) | 831.5 fps | 1005.1 fps |
| Car Chase | 517.6 fps | 897.2 fps |
| Manhattan 3.1.1 (1440p) | 1083.9 fps | 781.7 fps |
| T-Rex | 3144.1 fps | 4109.9 fps |
| ALU 2 | 2765.4 fps | 7102.5 fps |
| Driver Overhead 2 | 570.2 fps | 274.9 fps |
| API | Metal | OpenGL |
| OS | OS X | Windows |
| Tessellation | n/a | 7000.8 |
| Image Classification (SP) | 8094 | n/a |
| Image Segmentation (HP) | 24595 | n/a |
| Horizon Detection | 25190 | n/a |
| Face Detection (HP) | 38337 | n/a |
| Pose Estimation (Q) | 111021 | n/a |
| Text Classification (SP) | 2786 | n/a |
| Machine Translation (HP) | 4339 | n/a |
| Object Detection (SP) | 7668 | n/a |
| Depth Estimation (Q) | 37488 | n/a |
| Style Transfer (SP) | 244483 | n/a |
| Framework | Core ML | n/a |
| Backend | GPU | n/a |
Blender
Tech Specs
Specs & test comparisonGeneral Info
| Manufacturer | Apple | Nvidia |
| Build | Integrated | Discrete |
| Announced | March, 5, 2025 | March, 29, 2022 |
| Form Factor | PC use | PC use |
| Use case | Professional | Gaming |
| Segment | High-end | High-end |
| Architecture | Apple M GPU | Ampere |
| Codename | Custom | GA102 |
| Price | n/a | 1999 |
GPU
| GPU Frequency | n/a | 1560 MHz |
| Boost GPU Frequency | n/a | 1860 MHz |
| Shader Units | 10240 | 10752 |
| TMUs | 640 | 336 |
| ROPs | 320 | 112 |
| Execution Units | 1280 | 84 |
| Ray Tracing Cores | n/a | 84 |
| Tensor Cores | n/a | 336 |
| L1 cache | n/a | 128 KB/core |
| L2 cache | n/a | 6000 KB/core |
| Instructions per Clock | 2 IPC | 2 IPC |
Performance
| Pixel Fill Rate | n/a | 208 Gpix/sec |
| Texture Fill Rate | n/a | 625 GTex/s |
| FLOPS | n/a | 40 TFLOPS |
Process
| Interface | Custom | PCIe 4.0 x16 |
| TGP | 140 W | 450 W |
| Process | TSMC | Samsung |
| Process tech. | 3 nm | 8 nm |
| Max temperature | 100 °C | n/a |
| Transistor Size | n/a | 628 mm² |
| Transistor count | 184 billion | 28 billion |
| Transistor density | n/a | 44.59 MTr/mm² |
Memory
| Interface | System Shared | GDDR6X |
| Memory Frequency | 6400 MHz | 1313 MHz |
| Bus | 1024 bit | 384 bit |
| ECC support | No | No |
| Effective Memory Speed | n/a | 21000 MB/s |
| Memory Size | n/a | 24 GB |
Max Throughput
API
| DirectX | n/a | 12 |
| Vulkan | n/a | 1.3 |
| OpenGL | n/a | 4.6 |
| CUDA | n/a | 8.6 |
| DisplayPort | n/a | 1.4a |
| Ray Tracing | Yes | Yes |
| DLSS | No | DLSS 2 |
Results
After analyzing the data, we consider Nvidia GeForce RTX 3090 Ti the clear winner in this comparison.
Survey
Total Votes: 0
Forza Horizon 5
The Witcher 3
Counter-Strike 2
Far Cry 6
Hogwarts Legacy
COD Modern Warfare 3
Ghost of Tsushima
Cyberpunk 2077
Shadow of the Tomb Raider