Geekbench ML Score

2989
Core ML CPU Inference Score
Geekbench ML 0.6.0 for iOS AArch64

Result Information

Upload Date August 01 2024 07:05 PM
Views 1

System Information

System Information
Operating System iOS 17.6
Model iPad Pro (11-inch 3rd generation)
Model IDiPad13,7
Motherboard J518xAP
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 15.33 GB
Inference Information
Framework Core ML
Backend CPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 1294
242.2 IPS
 
Image Classification (F16)
100% 1675
313.4 IPS
 
Image Classification (I8)
99% 1675
313.5 IPS
 
Image Segmentation (F32)
100% 593
9.90 IPS
 
Image Segmentation (F16)
100% 591
9.88 IPS
 
Image Segmentation (I8)
100% 593
9.90 IPS
 
Pose Estimation (F32)
100% 5648
6.84 IPS
 
Pose Estimation (F16)
100% 12947
15.7 IPS
 
Pose Estimation (I8)
100% 12957
15.7 IPS
 
Object Detection (F32)
100% 1263
94.3 IPS
 
Object Detection (F16)
100% 1738
129.8 IPS
 
Object Detection (I8)
97% 1256
93.8 IPS
 
Face Detection (F32)
100% 2383
28.3 IPS
 
Face Detection (F16)
97% 4033
48.0 IPS
 
Face Detection (I8)
97% 4039
48.0 IPS
 
Depth Estimation (F32)
100% 5202
40.3 IPS
 
Depth Estimation (F16)
100% 8156
63.3 IPS
 
Depth Estimation (I8)
99% 8169
63.4 IPS
 
Style Transfer (F32)
100% 15638
20.6 IPS
 
Style Transfer (F16)
100% 24386
32.1 IPS
 
Style Transfer (I8)
100% 24385
32.1 IPS
 
Image Super-Resolution (F32)
100% 3498
124.9 IPS
 
Image Super-Resolution (F16)
100% 4509
161.0 IPS
 
Image Super-Resolution (I8)
100% 4542
162.2 IPS
 
Text Classification (F32)
100% 2104
3.02 KIPS
 
Text Classification (F16)
100% 3342
4.80 KIPS
 
Text Classification (I8)
96% 3300
4.74 KIPS
 
Machine Translation (F32)
100% 723
13.3 IPS
 
Machine Translation (F16)
100% 865
15.9 IPS
 
Machine Translation (I8)
99% 721
13.3 IPS
 

Compare

Set Baseline Geekbench 6 now available