Geekbench AI Score

14818
Single Precision Score
15747
Half Precision Score
12701
Quantized Score
Geekbench AI 1.2.0 for macOS AArch64

Result Information

User unixwizard
Upload Date December 10 2024 02:02 AM
Views 9

System Information

AI Information
Framework Core ML
Backend GPU
Device Apple M2 Max
System Information
Operating System macOS 15.1.1 (Build 24B91)
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 3.67 GHz
Cluster 18 Cores
Cluster 24 Cores
Memory Information
Size 32.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 6971
1.30 KIPS
 
Image Classification (HP)
100% 6855
1.27 KIPS
 
Image Classification (Q)
97% 6032
1.13 KIPS
 
Image Segmentation (SP)
100% 13740
222.7 IPS
 
Image Segmentation (HP)
100% 13514
219.1 IPS
 
Image Segmentation (Q)
99% 13476
218.5 IPS
 
Pose Estimation (SP)
100% 83240
97.1 IPS
 
Pose Estimation (HP)
100% 86570
101.0 IPS
 
Pose Estimation (Q)
95% 82058
96.2 IPS
 
Object Detection (SP)
100% 6458
512.3 IPS
 
Object Detection (HP)
100% 6726
533.5 IPS
 
Object Detection (Q)
91% 5861
468.3 IPS
 
Face Detection (SP)
100% 27973
332.4 IPS
 
Face Detection (HP)
100% 28675
340.7 IPS
 
Face Detection (Q)
100% 21072
250.4 IPS
 
Depth Estimation (SP)
100% 34717
267.5 IPS
 
Depth Estimation (HP)
99% 34761
267.8 IPS
 
Depth Estimation (Q)
85% 10668
83.4 IPS
 
Style Transfer (SP)
100% 132972
170.9 IPS
 
Style Transfer (HP)
100% 149570
192.3 IPS
 
Style Transfer (Q)
96% 140590
181.4 IPS
 
Image Super-Resolution (SP)
100% 18593
686.6 IPS
 
Image Super-Resolution (HP)
100% 19418
717.0 IPS
 
Image Super-Resolution (Q)
99% 13700
505.9 IPS
 
Text Classification (SP)
100% 1909
2.55 KIPS
 
Text Classification (HP)
100% 2847
3.80 KIPS
 
Text Classification (Q)
93% 2979
4.00 KIPS
 
Machine Translation (SP)
100% 2164
37.3 IPS
 
Machine Translation (HP)
100% 2110
36.3 IPS
 
Machine Translation (Q)
98% 2168
37.5 IPS
 

Compare

Set Baseline Geekbench 6 now available