Geekbench AI Score

705
Single Precision Score
973
Half Precision Score
822
Quantized Score
Geekbench AI 1.4.0 for Android AArch64

Result Information

Upload Date June 30 2025 09:06 PM
Views 4

System Information

AI Information
Framework TensorFlow Lite
Backend GPU
Device MediaTek Dimensity 8020
System Information
Operating System Android 15
Model Motorola Edge 40
Model IDmotorola motorola edge 40
Motherboard lyriq
CPU Information
Name ARM ARMv8
Topology 1 Processor, 8 Cores
Identifier ARM implementer 65 architecture 8 variant 1 part 3393 revision 0
Base Frequency 2.00 GHz
Cluster 14 Cores @ 2.00 GHz
Cluster 24 Cores @ 2.60 GHz
Memory Information
Size 7.28 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 429
79.8 IPS
 
Image Classification (HP)
100% 685
127.4 IPS
 
Image Classification (Q)
97% 633
118.2 IPS
 
Image Segmentation (SP)
100% 797
12.9 IPS
 
Image Segmentation (HP)
100% 1079
17.5 IPS
 
Image Segmentation (Q)
98% 1018
16.6 IPS
 
Pose Estimation (SP)
100% 3693
4.31 IPS
 
Pose Estimation (HP)
100% 6560
7.65 IPS
 
Pose Estimation (Q)
95% 6306
7.39 IPS
 
Object Detection (SP)
100% 331
26.3 IPS
 
Object Detection (HP)
99% 518
41.1 IPS
 
Object Detection (Q)
83% 494
39.9 IPS
 
Face Detection (SP)
100% 1216
14.4 IPS
 
Face Detection (HP)
100% 1920
22.8 IPS
 
Face Detection (Q)
96% 1746
20.8 IPS
 
Depth Estimation (SP)
100% 1959
15.1 IPS
 
Depth Estimation (HP)
98% 3188
24.6 IPS
 
Depth Estimation (Q)
62% 2342
22.2 IPS
 
Style Transfer (SP)
100% 5670
7.29 IPS
 
Style Transfer (HP)
100% 11947
15.4 IPS
 
Style Transfer (Q)
98% 11811
15.2 IPS
 
Image Super-Resolution (SP)
100% 1005
37.1 IPS
 
Image Super-Resolution (HP)
100% 1263
46.6 IPS
 
Image Super-Resolution (Q)
97% 1188
44.0 IPS
 
Text Classification (SP)
35% 14
127.0 IPS
 
Text Classification (HP)
29% 9
159.7 IPS
 
Text Classification (Q)
32% 14
182.3 IPS
 
Machine Translation (SP)
100% 387
6.66 IPS
 
Machine Translation (HP)
100% 367
6.32 IPS
 
Machine Translation (Q)
43% 88
5.12 IPS
 

Compare

Set Baseline Geekbench 6 now available