Geekbench ML Score

3005
Core ML CPU Inference Score
Geekbench ML 0.6.0 for iOS AArch64

Result Information

Upload Date May 29 2024 08:10 AM
Views 1

System Information

System Information
Operating System iOS 17.5.1
Model iPad Pro (12.9-inch 5th generation)
Model IDiPad13,10
Motherboard J523AP
CPU Information
Name Apple M1
Topology 1 Processor, 8 Cores
Identifier ARM
Base Frequency 3.19 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 7.31 GB
Inference Information
Framework Core ML
Backend CPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 1289
241.3 IPS
 
Image Classification (F16)
100% 1673
313.1 IPS
 
Image Classification (I8)
99% 1672
312.9 IPS
 
Image Segmentation (F32)
100% 596
9.95 IPS
 
Image Segmentation (F16)
100% 615
10.3 IPS
 
Image Segmentation (I8)
100% 591
9.87 IPS
 
Pose Estimation (F32)
100% 5656
6.85 IPS
 
Pose Estimation (F16)
100% 12941
15.7 IPS
 
Pose Estimation (I8)
100% 12947
15.7 IPS
 
Object Detection (F32)
100% 1260
94.0 IPS
 
Object Detection (F16)
100% 1731
129.2 IPS
 
Object Detection (I8)
97% 1271
94.9 IPS
 
Face Detection (F32)
100% 2393
28.5 IPS
 
Face Detection (F16)
97% 3994
47.5 IPS
 
Face Detection (I8)
97% 4007
47.6 IPS
 
Depth Estimation (F32)
100% 5188
40.2 IPS
 
Depth Estimation (F16)
100% 8118
63.0 IPS
 
Depth Estimation (I8)
99% 8124
63.0 IPS
 
Style Transfer (F32)
100% 15617
20.5 IPS
 
Style Transfer (F16)
100% 24271
31.9 IPS
 
Style Transfer (I8)
100% 24272
31.9 IPS
 
Image Super-Resolution (F32)
100% 3503
125.1 IPS
 
Image Super-Resolution (F16)
100% 4800
171.4 IPS
 
Image Super-Resolution (I8)
100% 4594
164.1 IPS
 
Text Classification (F32)
100% 2122
3.05 KIPS
 
Text Classification (F16)
100% 3427
4.92 KIPS
 
Text Classification (I8)
96% 3409
4.90 KIPS
 
Machine Translation (F32)
100% 734
13.5 IPS
 
Machine Translation (F16)
100% 861
15.8 IPS
 
Machine Translation (I8)
99% 725
13.3 IPS
 

Compare

Set Baseline Geekbench 6 now available