Geekbench ML Score

1746
Core ML Neural Engine Inference Score
Geekbench ML 0.6.0 for iOS AArch64

Result Information

Upload Date January 20 2024 08:08 PM
Views 1

System Information

System Information
Operating System iOS 17.2.1
Model iPhone 11
Model IDiPhone12,1
Motherboard N104AP
CPU Information
Name Apple A13 Bionic
Topology 1 Processor, 6 Cores
Identifier ARM
Base Frequency 2.66 GHz
Cluster 12 Cores
Cluster 24 Cores
L1 Instruction Cache96.0 KB x 1
L1 Data Cache48.0 KB x 1
L2 Cache4.00 MB x 1
Memory Information
Size 3.76 GB
Inference Information
Framework Core ML
Backend Neural Engine
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 273
51.1 IPS
 
Image Classification (F16)
100% 2602
486.9 IPS
 
Image Classification (I8)
99% 2602
486.9 IPS
 
Image Segmentation (F32)
100% 233
3.89 IPS
 
Image Segmentation (F16)
100% 331
5.53 IPS
 
Image Segmentation (I8)
100% 330
5.52 IPS
 
Pose Estimation (F32)
100% 2255
2.73 IPS
 
Pose Estimation (F16)
100% 26247
31.8 IPS
 
Pose Estimation (I8)
100% 25946
31.4 IPS
 
Object Detection (F32)
100% 408
30.5 IPS
 
Object Detection (F16)
100% 1171
87.4 IPS
 
Object Detection (I8)
97% 475
35.5 IPS
 
Face Detection (F32)
100% 2444
29.1 IPS
 
Face Detection (F16)
97% 4820
57.3 IPS
 
Face Detection (I8)
96% 4883
58.1 IPS
 
Depth Estimation (F32)
100% 2464
19.1 IPS
 
Depth Estimation (F16)
100% 21495
166.7 IPS
 
Depth Estimation (I8)
99% 21083
163.5 IPS
 
Style Transfer (F32)
24% 5897
7.76 IPS
 
Style Transfer (F16)
100% 17902
23.5 IPS
 
Style Transfer (I8)
100% 18008
23.7 IPS
 
Image Super-Resolution (F32)
100% 1375
49.1 IPS
 
Image Super-Resolution (F16)
100% 5019
179.2 IPS
 
Image Super-Resolution (I8)
100% 5137
183.5 IPS
 
Text Classification (F32)
100% 241
346.1 IPS
 
Text Classification (F16)
100% 310
444.8 IPS
 
Text Classification (I8)
96% 310
446.0 IPS
 
Machine Translation (F32)
100% 224
4.13 IPS
 
Machine Translation (F16)
100% 222
4.08 IPS
 
Machine Translation (I8)
99% 227
4.18 IPS
 

Compare

Set Baseline Geekbench 6 now available