Geekbench ML Score

2747
Core ML GPU Inference Score
Geekbench ML 0.6.0 for iOS AArch64

Result Information

User M91
Upload Date May 10 2024 12:25 PM
Views 2

System Information

System Information
Operating System iOS 17.5
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 GPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 1230
230.2 IPS
 
Image Classification (F16)
100% 1204
225.4 IPS
 
Image Classification (I8)
99% 1544
289.0 IPS
 
Image Segmentation (F32)
100% 1161
19.4 IPS
 
Image Segmentation (F16)
100% 1215
20.3 IPS
 
Image Segmentation (I8)
100% 1206
20.1 IPS
 
Pose Estimation (F32)
100% 11407
13.8 IPS
 
Pose Estimation (F16)
100% 12170
14.7 IPS
 
Pose Estimation (I8)
100% 12007
14.5 IPS
 
Object Detection (F32)
100% 1551
115.8 IPS
 
Object Detection (F16)
100% 1430
106.7 IPS
 
Object Detection (I8)
97% 1655
123.6 IPS
 
Face Detection (F32)
100% 5023
59.7 IPS
 
Face Detection (F16)
98% 5338
63.5 IPS
 
Face Detection (I8)
97% 5425
64.5 IPS
 
Depth Estimation (F32)
100% 8652
67.1 IPS
 
Depth Estimation (F16)
100% 9124
70.8 IPS
 
Depth Estimation (I8)
37% 8538
66.2 IPS
 
Style Transfer (F32)
100% 21081
27.7 IPS
 
Style Transfer (F16)
100% 23632
31.1 IPS
 
Style Transfer (I8)
100% 23366
30.7 IPS
 
Image Super-Resolution (F32)
100% 4324
154.4 IPS
 
Image Super-Resolution (F16)
100% 4411
157.5 IPS
 
Image Super-Resolution (I8)
100% 4325
154.5 IPS
 
Text Classification (F32)
100% 367
527.8 IPS
 
Text Classification (F16)
100% 354
509.4 IPS
 
Text Classification (I8)
96% 236
339.1 IPS
 
Machine Translation (F32)
100% 585
10.8 IPS
 
Machine Translation (F16)
100% 630
11.6 IPS
 
Machine Translation (I8)
99% 584
10.7 IPS
 

Compare

Set Baseline Geekbench 6 now available