Geekbench AI Score

13490
Single Precision Score
14563
Half Precision Score
11462
Quantized Score
Geekbench AI 1.0.0 for macOS AArch64

Result Information

User KeviNW000
Upload Date November 10 2024 08:39 PM
Views 4

System Information

AI Information
Framework Core ML
Backend GPU
Device Apple M2 Max
System Information
Operating System macOS 15.2 (Build 24C5073e)
Model Mac Studio (2023)
Model IDMac14,13
Motherboard Mac14,13
CPU Information
Name Apple M2 Max
Topology 1 Processor, 12 Cores
Identifier Apple M2 Max
Base Frequency 3.47 GHz
Cluster 18 Cores
Cluster 24 Cores
Memory Information
Size 32.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 6269
1.17 KIPS
 
Image Classification (HP)
100% 6522
1.21 KIPS
 
Image Classification (Q)
99% 6472
1.21 KIPS
 
Image Segmentation (SP)
100% 12085
195.9 IPS
 
Image Segmentation (HP)
100% 11819
191.6 IPS
 
Image Segmentation (Q)
100% 11882
192.6 IPS
 
Pose Estimation (SP)
100% 62102
72.5 IPS
 
Pose Estimation (HP)
100% 65223
76.1 IPS
 
Pose Estimation (Q)
97% 58613
68.6 IPS
 
Object Detection (SP)
100% 6092
483.2 IPS
 
Object Detection (HP)
100% 6478
513.8 IPS
 
Object Detection (Q)
91% 5748
459.3 IPS
 
Face Detection (SP)
100% 25322
300.9 IPS
 
Face Detection (HP)
100% 25968
308.6 IPS
 
Face Detection (Q)
100% 18669
221.8 IPS
 
Depth Estimation (SP)
100% 29069
224.0 IPS
 
Depth Estimation (HP)
99% 30086
231.8 IPS
 
Depth Estimation (Q)
97% 6067
46.9 IPS
 
Style Transfer (SP)
100% 114682
147.4 IPS
 
Style Transfer (HP)
100% 128387
165.0 IPS
 
Style Transfer (Q)
100% 129193
166.1 IPS
 
Image Super-Resolution (SP)
100% 20188
745.4 IPS
 
Image Super-Resolution (HP)
100% 20711
764.8 IPS
 
Image Super-Resolution (Q)
98% 16556
613.4 IPS
 
Text Classification (SP)
100% 2288
3.05 KIPS
 
Text Classification (HP)
100% 3864
5.16 KIPS
 
Text Classification (Q)
93% 3851
5.17 KIPS
 
Machine Translation (SP)
100% 1786
30.8 IPS
 
Machine Translation (HP)
100% 1642
28.3 IPS
 
Machine Translation (Q)
98% 1620
28.0 IPS
 

Compare

Set Baseline Geekbench 6 now available