Geekbench AI Score

1235
Single Precision Score
2007
Half Precision Score
2207
Quantized Score
Geekbench AI 1.2.0 for Android AArch64

Result Information

User Mystline
Upload Date January 31 2025 08:25 PM
Views 16

System Information

AI Information
Framework TensorFlow Lite
Backend GPU
Device Qualcomm ARMv8
System Information
Operating System Android 15
Model samsung SM-S938W
Model IDsamsung SM-S938W
Motherboard sun
Governorwalt
CPU Information
Name Qualcomm ARMv8
Topology 1 Processor, 8 Cores
Identifier ARM implementer 81 architecture 8 variant 3 part 1 revision 4
Base Frequency 3.53 GHz
Cluster 16 Cores @ 3.53 GHz
Cluster 22 Cores @ 4.47 GHz
Memory Information
Size 10.86 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 724
134.6 IPS
 
Image Classification (HP)
100% 827
153.9 IPS
 
Image Classification (Q)
99% 1752
326.8 IPS
 
Image Segmentation (SP)
100% 1598
25.9 IPS
 
Image Segmentation (HP)
100% 1937
31.4 IPS
 
Image Segmentation (Q)
98% 2561
41.6 IPS
 
Pose Estimation (SP)
100% 1641
1.91 IPS
 
Pose Estimation (HP)
100% 2471
2.88 IPS
 
Pose Estimation (Q)
96% 2463
2.89 IPS
 
Object Detection (SP)
100% 505
40.0 IPS
 
Object Detection (HP)
100% 1191
94.5 IPS
 
Object Detection (Q)
86% 1157
93.0 IPS
 
Face Detection (SP)
100% 2779
33.0 IPS
 
Face Detection (HP)
100% 4088
48.6 IPS
 
Face Detection (Q)
97% 4396
52.4 IPS
 
Depth Estimation (SP)
100% 1484
11.4 IPS
 
Depth Estimation (HP)
99% 4064
31.4 IPS
 
Depth Estimation (Q)
63% 3450
32.1 IPS
 
Style Transfer (SP)
100% 5094
6.55 IPS
 
Style Transfer (HP)
100% 10784
13.9 IPS
 
Style Transfer (Q)
98% 10816
13.9 IPS
 
Image Super-Resolution (SP)
100% 629
23.2 IPS
 
Image Super-Resolution (HP)
100% 1978
73.0 IPS
 
Image Super-Resolution (Q)
97% 1909
70.7 IPS
 
Text Classification (SP)
100% 655
874.5 IPS
 
Text Classification (HP)
100% 648
865.2 IPS
 
Text Classification (Q)
91% 1035
1.39 KIPS
 
Machine Translation (SP)
100% 997
17.2 IPS
 
Machine Translation (HP)
100% 984
17.0 IPS
 
Machine Translation (Q)
57% 663
16.2 IPS
 

Compare

Set Baseline Geekbench 6 now available