Geekbench ML Score

2189
Core ML CPU Inference Score
Geekbench ML 0.6.0 for macOS AArch64

Result Information

Upload Date April 26 2024 06:53 AM
Views 2

System Information

System Information
Operating System macOS 14.4.1 (Build 23E224)
Model MacBook Air (2022)
Model IDMac14,2
Motherboard Mac14,2
CPU Information
Name Apple M2
Topology 1 Processor, 8 Cores
Identifier Apple M2
Base Frequency 1.95 GHz
Cluster 14 Cores
Cluster 24 Cores
L1 Instruction Cache128 KB x 1
L1 Data Cache64.0 KB x 1
L2 Cache4.00 MB x 1
Memory Information
Size 8.00 GB
Inference Information
Framework Core ML
Backend CPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 916
171.4 IPS
 
Image Classification (F16)
100% 1121
209.8 IPS
 
Image Classification (I8)
99% 1129
211.3 IPS
 
Image Segmentation (F32)
100% 573
9.57 IPS
 
Image Segmentation (F16)
100% 555
9.27 IPS
 
Image Segmentation (I8)
100% 555
9.27 IPS
 
Pose Estimation (F32)
100% 3548
4.30 IPS
 
Pose Estimation (F16)
100% 8738
10.6 IPS
 
Pose Estimation (I8)
100% 8727
10.6 IPS
 
Object Detection (F32)
100% 935
69.8 IPS
 
Object Detection (F16)
100% 1226
91.6 IPS
 
Object Detection (I8)
97% 929
69.3 IPS
 
Face Detection (F32)
100% 1863
22.2 IPS
 
Face Detection (F16)
97% 2641
31.4 IPS
 
Face Detection (I8)
97% 2591
30.8 IPS
 
Depth Estimation (F32)
100% 3389
26.3 IPS
 
Depth Estimation (F16)
100% 5292
41.0 IPS
 
Depth Estimation (I8)
99% 5276
40.9 IPS
 
Style Transfer (F32)
100% 10482
13.8 IPS
 
Style Transfer (F16)
100% 16174
21.3 IPS
 
Style Transfer (I8)
100% 16161
21.3 IPS
 
Image Super-Resolution (F32)
100% 2414
86.2 IPS
 
Image Super-Resolution (F16)
100% 3219
115.0 IPS
 
Image Super-Resolution (I8)
100% 3221
115.0 IPS
 
Text Classification (F32)
100% 1382
1.99 KIPS
 
Text Classification (F16)
100% 1931
2.77 KIPS
 
Text Classification (I8)
96% 1972
2.83 KIPS
 
Machine Translation (F32)
100% 824
15.2 IPS
 
Machine Translation (F16)
100% 933
17.2 IPS
 
Machine Translation (I8)
99% 822
15.1 IPS
 

Compare

Set Baseline Geekbench 6 now available