Geekbench AI Score

12002
Single Precision Score
13366
Half Precision Score
11944
Quantized Score
Geekbench AI 1.1.0 for macOS AArch64

Result Information

User Wolfgang1756
Upload Date December 20 2024 09:58 PM
Views 3

System Information

AI Information
Framework Core ML
Backend GPU
Device Apple M1 Ultra
System Information
Operating System macOS 15.2 (Build 24C101)
Model Mac Studio
Model IDMac13,2
Motherboard Mac13,2
CPU Information
Name Apple M1 Ultra
Topology 1 Processor, 20 Cores
Identifier Apple M1 Ultra
Base Frequency 3.16 GHz
Cluster 116 Cores
Cluster 24 Cores
Memory Information
Size 128.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 4941
918.9 IPS
 
Image Classification (HP)
100% 4937
918.0 IPS
 
Image Classification (Q)
97% 4162
776.8 IPS
 
Image Segmentation (SP)
100% 12245
198.5 IPS
 
Image Segmentation (HP)
100% 12370
200.5 IPS
 
Image Segmentation (Q)
99% 12473
202.2 IPS
 
Pose Estimation (SP)
100% 81885
95.5 IPS
 
Pose Estimation (HP)
100% 84453
98.5 IPS
 
Pose Estimation (Q)
95% 78013
91.4 IPS
 
Object Detection (SP)
100% 3937
312.3 IPS
 
Object Detection (HP)
100% 4671
370.5 IPS
 
Object Detection (Q)
91% 4454
355.9 IPS
 
Face Detection (SP)
100% 22646
269.1 IPS
 
Face Detection (HP)
100% 24910
296.0 IPS
 
Face Detection (Q)
100% 17789
211.4 IPS
 
Depth Estimation (SP)
100% 27912
215.0 IPS
 
Depth Estimation (HP)
99% 28481
219.4 IPS
 
Depth Estimation (Q)
85% 24130
188.6 IPS
 
Style Transfer (SP)
100% 135584
174.3 IPS
 
Style Transfer (HP)
100% 152573
196.1 IPS
 
Style Transfer (Q)
96% 141226
182.2 IPS
 
Image Super-Resolution (SP)
100% 16426
606.5 IPS
 
Image Super-Resolution (HP)
100% 16387
605.1 IPS
 
Image Super-Resolution (Q)
99% 13091
483.4 IPS
 
Text Classification (SP)
100% 1985
2.65 KIPS
 
Text Classification (HP)
100% 3876
5.17 KIPS
 
Text Classification (Q)
93% 3839
5.15 KIPS
 
Machine Translation (SP)
100% 1138
19.6 IPS
 
Machine Translation (HP)
100% 1099
18.9 IPS
 
Machine Translation (Q)
98% 1076
18.6 IPS
 

Compare

Set Baseline Geekbench 6 now available