Geekbench ML Score

504
TensorFlow Lite GPU Inference Score
Geekbench ML 0.5.0 for Android AArch64

Result Information

Upload Date October 23 2021 02:50 AM
Views 1

System Information

System Information
Operating System Android 10
Model samsung SM-N960N
Model IDsamsung SM-N960N
Motherboard exynos9810
Governorschedutil
CPU Information
Name ARMv8
Topology 1 Processor, 8 Cores
Identifier ARM implementer 83 architecture 8 variant 1 part 2 revision 0
Base Frequency 1.79 GHz
Cluster 14 Cores @ 1.79 GHz
Cluster 24 Cores @ 2.70 GHz
Memory Information
Size 7.45 GB
Inference Information
Framework TensorFlow Lite
Backend GPU

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 383
41.1 IPS
 
Image Classification (F16)
100% 634
68.0 IPS
 
Image Classification (I8)
81% 585
62.6 IPS
 
Image Segmentation (F32)
100% 779
4.81 IPS
 
Image Segmentation (F16)
100% 1758
10.9 IPS
 
Image Segmentation (I8)
89% 1685
10.4 IPS
 
Pose Estimation (F32)
100% 175
0.71 IPS
 
Pose Estimation (F16)
0% 604
2.44 IPS
 
Pose Estimation (I8)
0% 606
2.45 IPS
 
Object Detection (F32)
100% 170
5.97 IPS
 
Object Detection (F16)
99% 283
9.95 IPS
 
Object Detection (I8)
54% 238
8.36 IPS
 
Face Detection (F32)
100% 738
8.48 IPS
 
Face Detection (F16)
97% 1446
16.6 IPS
 
Face Detection (I8)
75% 1240
14.2 IPS
 
Text Classification (F32)
100% 314
124.2 IPS
 
Text Classification (F16)
98% 436
172.4 IPS
 
Machine Translation (F32)
100% 231
35.6 IPS
 
Machine Translation (F16)
100% 227
35.0 IPS
 

Compare

Set Baseline

Share

Geekbench ML now available for download