Geekbench AI Score

359
Single Precision Score
392
Half Precision Score
49
Quantized Score
Geekbench AI 1.1.0 for Android AArch64

Result Information

Upload Date September 23 2024 01:03 PM
Views 28

System Information

AI Information
Framework TensorFlow Lite
Backend NNAPI
Device Qualcomm Snapdragon 888
System Information
Operating System Android 14
Model Samsung Galaxy Z Fold3 5G
Model IDsamsung SM-F926B
Motherboard lahaina
Governorschedutil
CPU Information
Name ARM SM8350
Topology 1 Processor, 8 Cores
Identifier ARM implementer 65 architecture 8 variant 1 part 3396 revision 0
Base Frequency 1.80 GHz
Cluster 14 Cores @ 1.80 GHz
Cluster 23 Cores @ 2.42 GHz
Cluster 31 Core @ 2.84 GHz
Memory Information
Size 11.09 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 531
98.7 IPS
 
Image Classification (HP)
100% 974
181.2 IPS
 
Image Classification (Q)
94% 17
3.22 IPS
 
Image Segmentation (SP)
100% 878
14.2 IPS
 
Image Segmentation (HP)
100% 1294
21.0 IPS
 
Image Segmentation (Q)
98% 85
1.38 IPS
 
Pose Estimation (SP)
100% 1998
2.33 IPS
 
Pose Estimation (HP)
100% 6429
7.50 IPS
 
Pose Estimation (Q)
98% 248
0.29 IPS
 
Object Detection (SP)
100% 540
42.8 IPS
 
Object Detection (HP)
99% 904
71.7 IPS
 
Object Detection (Q)
2% 8
190.3 IPS
 
Face Detection (SP)
100% 1464
17.4 IPS
 
Face Detection (HP)
100% 81
0.96 IPS
 
Face Detection (Q)
97% 218
2.60 IPS
 
Depth Estimation (SP)
100% 183
1.41 IPS
 
Depth Estimation (HP)
99% 76
0.59 IPS
 
Depth Estimation (Q)
5% 32
55.0 IPS
 
Style Transfer (SP)
100% 694
0.89 IPS
 
Style Transfer (HP)
100% 136
0.17 IPS
 
Style Transfer (Q)
18% 9
0.52 IPS
 
Image Super-Resolution (SP)
100% 989
36.5 IPS
 
Image Super-Resolution (HP)
100% 3113
115.0 IPS
 
Image Super-Resolution (Q)
97% 255
9.45 IPS
 
Text Classification (SP)
100% 4
5.99 IPS
 
Text Classification (HP)
100% 52
70.0 IPS
 
Text Classification (Q)
93% 26
35.5 IPS
 
Machine Translation (SP)
100% 97
1.66 IPS
 
Machine Translation (HP)
100% 87
1.50 IPS
 
Machine Translation (Q)
57% 76
1.86 IPS
 

Compare

Set Baseline Geekbench 6 now available