Geekbench AI Score

11459
Single Precision Score
11437
Half Precision Score
10476
Quantized Score
Geekbench AI 1.0.0 Corporate for macOS AArch64

Result Information

Upload Date August 15 2024 03:38 AM
Views 17

System Information

AI Information
Framework Core ML
Backend GPU
Device Apple M2 Ultra
System Information
Operating System macOS 14.5 (Build 23F79)
Model Mac Studio (2023)
Model IDMac14,14
Motherboard Mac14,14
CPU Information
Name Apple M2 Ultra
Topology 1 Processor, 24 Cores
Identifier Apple M2 Ultra
Base Frequency 3.48 GHz
Cluster 116 Cores
Cluster 28 Cores
Memory Information
Size 64.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 1107
205.9 IPS
 
Image Classification (HP)
100% 1015
188.8 IPS
 
Image Classification (Q)
99% 983
183.4 IPS
 
Image Segmentation (SP)
100% 14948
242.3 IPS
 
Image Segmentation (HP)
100% 15245
247.1 IPS
 
Image Segmentation (Q)
100% 14579
236.3 IPS
 
Pose Estimation (SP)
100% 103646
120.9 IPS
 
Pose Estimation (HP)
100% 107655
125.6 IPS
 
Pose Estimation (Q)
96% 92507
108.4 IPS
 
Object Detection (SP)
100% 5808
460.7 IPS
 
Object Detection (HP)
100% 5844
463.5 IPS
 
Object Detection (Q)
92% 4781
381.6 IPS
 
Face Detection (SP)
100% 25542
303.5 IPS
 
Face Detection (HP)
100% 26256
312.0 IPS
 
Face Detection (Q)
100% 19417
230.7 IPS
 
Depth Estimation (SP)
100% 32261
248.6 IPS
 
Depth Estimation (HP)
99% 34030
262.2 IPS
 
Depth Estimation (Q)
97% 33919
262.3 IPS
 
Style Transfer (SP)
100% 189021
243.0 IPS
 
Style Transfer (HP)
100% 193913
249.3 IPS
 
Style Transfer (Q)
100% 188280
242.0 IPS
 
Image Super-Resolution (SP)
100% 23863
881.2 IPS
 
Image Super-Resolution (HP)
100% 24137
891.3 IPS
 
Image Super-Resolution (Q)
98% 17333
642.1 IPS
 
Text Classification (SP)
100% 522
697.0 IPS
 
Text Classification (HP)
100% 443
591.5 IPS
 
Text Classification (Q)
93% 553
742.3 IPS
 
Machine Translation (SP)
100% 2021
34.8 IPS
 
Machine Translation (HP)
96% 2125
36.7 IPS
 
Machine Translation (Q)
98% 2115
36.5 IPS
 

Compare

Set Baseline Geekbench 6 now available