Geekbench ML Score

3176
Core ML CPU Inference Score
Geekbench ML 0.6.0 for iOS AArch64

Result Information

Upload Date May 22 2024 04:11 PM
Views 1

System Information

System Information
Operating System iOS 17.4.1
Model iPad mini (6th generation)
Model IDiPad14,2
Motherboard J311AP
CPU Information
Name Apple A15 Bionic
Topology 1 Processor, 6 Cores
Identifier ARM
Base Frequency 2.92 GHz
Cluster 12 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 3.78 GB
Inference Information
Framework Core ML
Backend CPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 1338
250.3 IPS
 
Image Classification (F16)
100% 1658
310.2 IPS
 
Image Classification (I8)
99% 1653
309.4 IPS
 
Image Segmentation (F32)
100% 758
12.7 IPS
 
Image Segmentation (F16)
100% 744
12.4 IPS
 
Image Segmentation (I8)
100% 747
12.5 IPS
 
Pose Estimation (F32)
100% 5562
6.74 IPS
 
Pose Estimation (F16)
100% 12763
15.5 IPS
 
Pose Estimation (I8)
100% 12780
15.5 IPS
 
Object Detection (F32)
100% 1296
96.8 IPS
 
Object Detection (F16)
100% 1752
130.8 IPS
 
Object Detection (I8)
97% 1303
97.3 IPS
 
Face Detection (F32)
100% 2321
27.6 IPS
 
Face Detection (F16)
97% 3742
44.5 IPS
 
Face Detection (I8)
97% 3738
44.4 IPS
 
Depth Estimation (F32)
100% 4779
37.1 IPS
 
Depth Estimation (F16)
100% 7752
60.1 IPS
 
Depth Estimation (I8)
99% 7737
60.0 IPS
 
Style Transfer (F32)
100% 15371
20.2 IPS
 
Style Transfer (F16)
100% 23711
31.2 IPS
 
Style Transfer (I8)
100% 23919
31.5 IPS
 
Image Super-Resolution (F32)
100% 3545
126.6 IPS
 
Image Super-Resolution (F16)
100% 4717
168.5 IPS
 
Image Super-Resolution (I8)
100% 4546
162.3 IPS
 
Text Classification (F32)
100% 2168
3.12 KIPS
 
Text Classification (F16)
100% 3512
5.05 KIPS
 
Text Classification (I8)
96% 3531
5.07 KIPS
 
Machine Translation (F32)
100% 1104
20.3 IPS
 
Machine Translation (F16)
100% 1378
25.4 IPS
 
Machine Translation (I8)
99% 1092
20.1 IPS
 

Compare

Set Baseline Geekbench 6 now available