Geekbench ML Score

544
TensorFlow Lite GPU Inference Score
Geekbench ML 0.5.0 for Android AArch64

Result Information

Upload Date August 06 2022 05:30 AM
Views 1

System Information

System Information
Operating System Android 12
Model Khadas rk3588_s
Model IDKhadas rk3588_s
Motherboard rk30sdk
Governorschedutil
CPU Information
Name ARM ARMv8
Topology 1 Processor, 8 Cores
Identifier ARM implementer 65 architecture 8 variant 4 part 3339 revision 0
Base Frequency 1.80 GHz
Cluster 14 Cores @ 1.80 GHz
Cluster 22 Cores @ 2.26 GHz
Cluster 32 Cores @ 2.30 GHz
Memory Information
Size 7.49 GB
Inference Information
Framework TensorFlow Lite
Backend GPU

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 395
42.3 IPS
 
Image Classification (F16)
100% 660
70.7 IPS
 
Image Classification (I8)
81% 583
62.5 IPS
 
Image Segmentation (F32)
100% 875
5.40 IPS
 
Image Segmentation (F16)
100% 1330
8.22 IPS
 
Image Segmentation (I8)
89% 1352
8.35 IPS
 
Pose Estimation (F32)
100% 284
1.15 IPS
 
Pose Estimation (F16)
100% 722
2.92 IPS
 
Pose Estimation (I8)
80% 723
2.93 IPS
 
Object Detection (F32)
100% 279
9.81 IPS
 
Object Detection (F16)
99% 382
13.4 IPS
 
Object Detection (I8)
54% 383
13.4 IPS
 
Face Detection (F32)
100% 392
4.50 IPS
 
Face Detection (F16)
97% 1389
16.0 IPS
 
Face Detection (I8)
75% 1211
13.9 IPS
 
Text Classification (F32)
100% 524
207.1 IPS
 
Text Classification (F16)
98% 808
319.6 IPS
 
Machine Translation (F32)
100% 154
23.8 IPS
 
Machine Translation (F16)
100% 154
23.7 IPS
 

Compare

Set Baseline

Share

Geekbench ML now available for download