Apple M3 Ultra GPU (80-core) vs M2 Max GPU (38-core)
We performed a head-to-head comparison of the Apple M3 Ultra GPU (80-core) with 1280 pipelines and 10240 shaders against the 2 years and 2 months older M2 Max GPU (38-core) that utilizes 608 pipelines and 4864 shaders. Here you will find complete details about specs, efficiency, performance tests, and more.
Review
Gaming
Performance in DirectX, OpenCL, and Vulkan games
Workstation
Perf. in 3D modeling, video editing and rendering apps
AI/ML
Capabilities for machine learning and AI-related tasks
Energy Efficiency
Power consumption efficiency in different scenarios
DeviceFacts Final Score
Overall video card score
The "Energy Efficiency" metric has less impact on the DeviceFacts Score for desktop GPU.
Key differences
Reasons to consider the Apple M3 Ultra GPU (80-core)
- Performs significantly better (up to 2.1x) in 3DMark Steel Nomad Lite
- 2.1x higher maximum theoretical performance (28.3 vs 13.6 TFLOPS)
- Manufactured using a more efficient 3 nm process technology
- Has 2x higher memory bandwidth: 819.3 vs 409.6 GB/s
- Achieves 70% more points in the GeekBench 6 Compute test (147K vs 86K)
- Has 2.1x more shading units (10240 vs 4864)
Benchmarks
Graphics cards’ performance in recent benchmarking apps3D Mark
Steel Nomad Lite Score
| Solar Bay | 63597 | 30368 |
| Wild Life Extreme | 47814 | 24948 |
GeekBench 6 OpenCL
GB6 Compute Score
| Background Blur | 204.5 img/sec | 149 img/sec |
| Face Detection | 126 img/sec | 98.1 img/sec |
| Horizon Detection | 6.53 Gpixels/sec | 3.8 Gpixels/sec |
| Edge Detection | 10.7 Gpixels/sec | 6.88 Gpixels/sec |
| Gaussian Blur | 9.5 Gpixels/sec | 4.44 Gpixels/sec |
| Feature Matching | 1.05 Gpixels/sec | 0.73 Gpixels/sec |
| Stereo Matching | 560.3 Gpixels/sec | 264.9 Gpixels/sec |
| Particle Physics | 25267.8 FPS | 11144.2 FPS |
| API | OpenCL | OpenCL |
Cinebench 2024 GPU
Cinebench 2024 GPU
Blender
Blender GPU
Recent User Tests
Apple M3 Ultra GPU (80-core)
| Date | Benchmark | Result |
|---|---|---|
| 📘 2025-03-16 (Tyler) | Geekbench 6 OpenCL | 145706 |
| 📘 2025-03-16 (Tyler) | Cinebench 2024 | 19448 |
Apple M2 Max GPU (38-core)
No benchmark results yet
Artificial Intelligence Tests
Performance in machine learning and artificial intelligence tasksGeekBench 6 ML
GB6 ML Single Precision
GB6 ML Half Precision
GB6 ML Quantized
| Image Classification (SP) | 8094 | 6657 |
| Image Segmentation (HP) | 24595 | 13854 |
| Image Super Resolution (Q) | 25190 | 18642 |
| Face Detection (HP) | 38337 | 27965 |
| Pose Estimation (Q) | 111021 | 83144 |
| Text Classification (SP) | 2786 | 2377 |
| Machine Translation (HP) | 4339 | 2052 |
| Object Detection (SP) | 7668 | 6495 |
| Depth Estimation (Q) | 37488 | 31105 |
| Style Transfer (SP) | 244483 | 135686 |
| Framework | Core ML | Core ML |
| Backend | GPU | GPU |
Specifications
Technical specifications of Apple M3 Ultra GPU (80-core) and M2 Max GPU (38-core)General
| Vendor | Apple | Apple |
| Build | Integrated | Integrated |
| Released | March 5, 2025 | January 17, 2023 |
| Case | Desktop | Laptop |
| Purpose | Professional | Professional |
| Segment | High-end | High-end |
| Architecture | Apple M GPU | Apple M GPU |
| GPU Codename | Custom | - |
| Rival Equivalent | - GeForce RTX 5070 Ti | - GeForce RTX 3060 Laptop |
| Successor | - Apple M4 Ultra GPU (80-core) | - Apple M5 Max GPU (40-core) |
| Recommended CPU | - Apple M3 Ultra or above | - |
| Used in CPUs | - Apple M3 Ultra | - Apple M2 Max |
Desktop GPU rating (#29th place)
Laptop GPU ranking (#42nd place)
Graphics Processing Unit
| Base Clock | - | 450 MHz |
| Boost Clock | 1380 MHz | 1398 MHz |
| Shading Units | 10240 | 4864 |
| Texture Mapping Units (TMUs) | 640 | 304 |
| Render Output Units (ROPs) | 320 | 152 |
| Compute Units (Pipelines) | 1280 | 608 |
| Instructions Per Cycle | 2 IPC | 2 IPC |
Raw Performance
| Pixel Fill Rate | 442 GPixel/s | 212 GPixel/s |
| Texture Fill Rate | 883 GTexel/s | 425 GTexel/s |
FLOPS (FP32)
Physical
| Interface | Custom | Custom |
| TGP | 140 W | 70 W |
| Manufacturing | TSMC | TSMC |
| Fabrication Process | 3 nm | 5 nm |
| Transistor Count | 184 billion | 52 billion |
| Max. Temperature | 100°C | 94°C |
Memory
| Memory Type | System Shared | System Shared |
| Memory Clock | 6400 MHz | 6400 MHz |
| Effective Memory Speed | - | 12800 Mbps |
| Bus | 1024-bit | 512-bit |
| ECC | No | No |
Memory Bandwidth
API
| Ray Tracing | Yes | No |
| DLSS | No | No |
Cast your vote
Total votes: < 1