Head to head
Compare Graphics Cards
Pick up to 4. Winners are highlighted in green ▲.
1 selected
Key specs
Memory Type
System Shared
Boost Clock
1380 MHz
TGP
140 W
Architecture
Apple M GPU
Benchmarks
Rating
53
Gaming
51
Render
59
TFLOPS (FP32)
28.3
Geekbench OpenCL
147,719
Steel Nomad Lite Score
19,584
GB6 Compute Score
147,719
Cinebench 2024 GPU
19,448
Blender GPU
7371.98
GB6 ML Single Precision
21,356
GB6 ML Half Precision
23,530
GB6 ML Quantized
21,706
FLOPS (FP32)
28.3 TFLOPS
Memory Bandwidth
819.3 GB/s
3D Mark
Solar Bay
63,597
Wild Life Extreme
47,814
GeekBench 6 OpenCL
Background Blur
204.5 img/sec
Face Detection
126 img/sec
Horizon Detection
6.53 Gpixels/sec
Edge Detection
10.7 Gpixels/sec
Gaussian Blur
9.5 Gpixels/sec
Feature Matching
1.05 Gpixels/sec
Stereo Matching
560.3 Gpixels/sec
Particle Physics
25267.8 FPS
API
OpenCL
GeekBench 6 ML
Image Classification (SP)
8094
Image Segmentation (HP)
24,595
Image Super Resolution (Q)
25,190
Face Detection (HP)
38,337
Pose Estimation (Q)
111,021
Text Classification (SP)
2786
Machine Translation (HP)
4339
Object Detection (SP)
7668
Depth Estimation (Q)
37,488
Style Transfer (SP)
244,483
Framework
Core ML
Backend
GPU
General
Vendor
Apple
Build
Integrated
Released
March 5, 2025
Case
Desktop
Purpose
Professional
Segment
High-end
Architecture
Apple M GPU
GPU Codename
Custom
Nvidia Equivalent
- GeForce RTX 5070 Ti
Successor
- Apple M4 Ultra GPU (80-core)
Recommended CPU
- Apple M3 Ultra or above
Used in CPUs
- Apple M3 Ultra
Graphics Processing Unit
Boost Clock
1380 MHz
Shading Units
10,240
Texture Mapping Units (TMUs)
640
Render Output Units (ROPs)
320
Compute Units (Pipelines)
1280
Instructions Per Cycle
2 IPC
Raw Performance
Pixel Fill Rate
442 GPixel/s
Texture Fill Rate
883 GTexel/s
Physical
Interface
Custom
TGP
140 W
Manufacturing
TSMC
Fabrication Process
3 nm
Transistor Count
184 billion
Max. Temperature
100°C
Memory
Memory Type
System Shared
Memory Clock
6400 MHz
Bus
1024-bit
ECC
No
API
Ray Tracing
Yes
DLSS
No