Geekbench AI Score

1949
Single Precision Score
12643
Half Precision Score
12805
Quantized Score
Geekbench AI 1.0.0 for macOS AArch64

Result Information

User frank_m_h_jaeger
Upload Date August 18 2024 11:27 AM
Views 12

System Information

AI Information
Framework Core ML
Backend Neural Engine
Device Apple M1
System Information
Operating System macOS 14.6.1 (Build 23G93)
Model MacBook Air (Late 2020)
Model IDMacBookAir10,1
Motherboard MacBookAir10,1
CPU Information
Name Apple M1
Topology 1 Processor, 8 Cores
Identifier Apple M1
Base Frequency 3.20 GHz
Cluster 14 Cores
Cluster 24 Cores
Memory Information
Size 16.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 991
184.3 IPS
 
Image Classification (HP)
100% 6477
1.20 KIPS
 
Image Classification (Q)
99% 6568
1.23 KIPS
 
Image Segmentation (SP)
100% 655
10.6 IPS
 
Image Segmentation (HP)
100% 4229
68.6 IPS
 
Image Segmentation (Q)
100% 4268
69.2 IPS
 
Pose Estimation (SP)
100% 5724
6.68 IPS
 
Pose Estimation (HP)
100% 67808
79.1 IPS
 
Pose Estimation (Q)
94% 64240
75.3 IPS
 
Object Detection (SP)
100% 1080
85.7 IPS
 
Object Detection (HP)
99% 7166
570.0 IPS
 
Object Detection (Q)
93% 6666
531.9 IPS
 
Face Detection (SP)
100% 1857
22.1 IPS
 
Face Detection (HP)
100% 12481
148.3 IPS
 
Face Detection (Q)
100% 12773
151.8 IPS
 
Depth Estimation (SP)
100% 3477
26.8 IPS
 
Depth Estimation (HP)
99% 44890
345.8 IPS
 
Depth Estimation (Q)
97% 47322
365.9 IPS
 
Style Transfer (SP)
100% 9904
12.7 IPS
 
Style Transfer (HP)
100% 119171
153.2 IPS
 
Style Transfer (Q)
100% 124588
160.2 IPS
 
Image Super-Resolution (SP)
100% 2497
92.2 IPS
 
Image Super-Resolution (HP)
100% 18050
666.5 IPS
 
Image Super-Resolution (Q)
98% 20634
764.5 IPS
 
Text Classification (SP)
100% 904
1.21 KIPS
 
Text Classification (HP)
100% 2805
3.74 KIPS
 
Text Classification (Q)
93% 2762
3.71 KIPS
 
Machine Translation (SP)
100% 1366
23.5 IPS
 
Machine Translation (HP)
100% 2320
40.0 IPS
 
Machine Translation (Q)
98% 2302
39.8 IPS
 

Compare

Set Baseline Geekbench 6 now available