Geekbench AI Score

5788
Single Precision Score
36734
Half Precision Score
49050
Quantized Score
Geekbench AI 1.1.0 for macOS AArch64

Result Information

User Suginoko
Upload Date November 19 2024 05:35 PM
Views 3

System Information

AI Information
Framework Core ML
Backend Neural Engine
Device Apple M4 Pro
System Information
Operating System macOS 15.1 (Build 24B2082)
Model Mac16,8
Model IDMac16,8
Motherboard Mac16,8
CPU Information
Name Apple M4 Pro
Topology 1 Processor, 12 Cores
Identifier Apple M4 Pro
Base Frequency 4.51 GHz
Cluster 18 Cores
Cluster 24 Cores
Memory Information
Size 48.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 2088
388.2 IPS
 
Image Classification (HP)
100% 17567
3.27 KIPS
 
Image Classification (Q)
97% 22614
4.22 KIPS
 
Image Segmentation (SP)
100% 2507
40.6 IPS
 
Image Segmentation (HP)
100% 23222
376.5 IPS
 
Image Segmentation (Q)
99% 32064
519.8 IPS
 
Pose Estimation (SP)
100% 18880
22.0 IPS
 
Pose Estimation (HP)
100% 123477
144.1 IPS
 
Pose Estimation (Q)
95% 231329
271.1 IPS
 
Object Detection (SP)
100% 2482
196.9 IPS
 
Object Detection (HP)
100% 19872
1.58 KIPS
 
Object Detection (Q)
91% 24586
1.96 KIPS
 
Face Detection (SP)
100% 4596
54.6 IPS
 
Face Detection (HP)
100% 45039
535.2 IPS
 
Face Detection (Q)
100% 54442
646.9 IPS
 
Depth Estimation (SP)
100% 8746
67.4 IPS
 
Depth Estimation (HP)
99% 106368
819.5 IPS
 
Depth Estimation (Q)
85% 180224
1.41 KIPS
 
Style Transfer (SP)
100% 35563
45.7 IPS
 
Style Transfer (HP)
100% 237855
305.8 IPS
 
Style Transfer (Q)
96% 301635
389.2 IPS
 
Image Super-Resolution (SP)
100% 7038
259.9 IPS
 
Image Super-Resolution (HP)
100% 47032
1.74 KIPS
 
Image Super-Resolution (Q)
99% 78835
2.91 KIPS
 
Text Classification (SP)
100% 2958
3.95 KIPS
 
Text Classification (HP)
100% 5029
6.71 KIPS
 
Text Classification (Q)
93% 5157
6.92 KIPS
 
Machine Translation (SP)
100% 5781
99.6 IPS
 
Machine Translation (HP)
100% 16587
285.7 IPS
 
Machine Translation (Q)
98% 16247
280.8 IPS
 

Compare

Set Baseline Geekbench 6 now available