Geekbench ML Score

3455
Core ML CPU Inference Score
Geekbench ML 0.6.0 for iOS AArch64

Result Information

Upload Date September 02 2024 09:24 PM
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% 1394
260.8 IPS
 
Image Classification (F16)
100% 1802
337.2 IPS
 
Image Classification (I8)
99% 1810
338.8 IPS
 
Image Segmentation (F32)
100% 762
12.7 IPS
 
Image Segmentation (F16)
100% 771
12.9 IPS
 
Image Segmentation (I8)
100% 767
12.8 IPS
 
Pose Estimation (F32)
100% 6010
7.28 IPS
 
Pose Estimation (F16)
100% 13634
16.5 IPS
 
Pose Estimation (I8)
100% 13661
16.5 IPS
 
Object Detection (F32)
100% 1404
104.8 IPS
 
Object Detection (F16)
100% 1916
143.0 IPS
 
Object Detection (I8)
97% 1397
104.3 IPS
 
Face Detection (F32)
100% 2648
31.5 IPS
 
Face Detection (F16)
97% 4264
50.7 IPS
 
Face Detection (I8)
97% 4260
50.7 IPS
 
Depth Estimation (F32)
100% 5384
41.8 IPS
 
Depth Estimation (F16)
100% 8345
64.7 IPS
 
Depth Estimation (I8)
99% 8325
64.6 IPS
 
Style Transfer (F32)
100% 16733
22.0 IPS
 
Style Transfer (F16)
100% 25571
33.6 IPS
 
Style Transfer (I8)
100% 25517
33.6 IPS
 
Image Super-Resolution (F32)
100% 3964
141.6 IPS
 
Image Super-Resolution (F16)
100% 4980
177.9 IPS
 
Image Super-Resolution (I8)
100% 5001
178.6 IPS
 
Text Classification (F32)
100% 2377
3.42 KIPS
 
Text Classification (F16)
100% 4045
5.81 KIPS
 
Text Classification (I8)
96% 4005
5.76 KIPS
 
Machine Translation (F32)
100% 1245
22.9 IPS
 
Machine Translation (F16)
100% 1471
27.1 IPS
 
Machine Translation (I8)
99% 1210
22.3 IPS
 

Compare

Set Baseline Geekbench 6 now available