Geekbench ML Score

3491
Core ML CPU Inference Score
Geekbench ML 0.6.0 for iOS AArch64

Result Information

Upload Date September 08 2024 04:46 AM
Views 1

System Information

System Information
Operating System iOS 17.5.1
Model iPhone 14 Pro Max
Model IDiPhone15,3
Motherboard D74AP
CPU Information
Name Apple A16 Bionic
Topology 1 Processor, 6 Cores
Identifier ARM
Base Frequency 3.46 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 5.48 GB
Inference Information
Framework Core ML
Backend CPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 1437
268.9 IPS
 
Image Classification (F16)
100% 1820
340.6 IPS
 
Image Classification (I8)
99% 1853
346.7 IPS
 
Image Segmentation (F32)
100% 761
12.7 IPS
 
Image Segmentation (F16)
100% 766
12.8 IPS
 
Image Segmentation (I8)
100% 758
12.7 IPS
 
Pose Estimation (F32)
100% 6211
7.52 IPS
 
Pose Estimation (F16)
100% 13930
16.9 IPS
 
Pose Estimation (I8)
100% 13859
16.8 IPS
 
Object Detection (F32)
100% 1436
107.2 IPS
 
Object Detection (F16)
100% 1971
147.2 IPS
 
Object Detection (I8)
97% 1425
106.4 IPS
 
Face Detection (F32)
100% 2705
32.2 IPS
 
Face Detection (F16)
97% 4322
51.4 IPS
 
Face Detection (I8)
97% 4324
51.4 IPS
 
Depth Estimation (F32)
100% 5516
42.8 IPS
 
Depth Estimation (F16)
100% 8521
66.1 IPS
 
Depth Estimation (I8)
99% 8528
66.1 IPS
 
Style Transfer (F32)
100% 17045
22.4 IPS
 
Style Transfer (F16)
100% 26071
34.3 IPS
 
Style Transfer (I8)
100% 25908
34.1 IPS
 
Image Super-Resolution (F32)
100% 3999
142.8 IPS
 
Image Super-Resolution (F16)
100% 4941
176.5 IPS
 
Image Super-Resolution (I8)
100% 5193
185.5 IPS
 
Text Classification (F32)
100% 2361
3.39 KIPS
 
Text Classification (F16)
100% 3875
5.57 KIPS
 
Text Classification (I8)
96% 3926
5.64 KIPS
 
Machine Translation (F32)
100% 1220
22.4 IPS
 
Machine Translation (F16)
100% 1519
28.0 IPS
 
Machine Translation (I8)
99% 1187
21.9 IPS
 

Compare

Set Baseline Geekbench 6 now available