Geekbench AI Score

1922
Single Precision Score
1770
Half Precision Score
3047
Quantized Score
Geekbench AI 1.5.0 for Android AArch64

Result Information

Upload Date November 13 2025 08:26 AM
Views 1

System Information

AI Information
Framework TensorFlow Lite
Backend CPU
Device Qualcomm ARMv8
System Information
Operating System Android 16
Model Xiaomi 25010PN30G
Model IDXiaomi 25010PN30G
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 @ 4.32 GHz
Memory Information
Size 14.76 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 1682
312.8 IPS
 
Image Classification (HP)
100% 1671
310.8 IPS
 
Image Classification (Q)
100% 3759
699.0 IPS
 
Image Segmentation (SP)
100% 2051
33.3 IPS
 
Image Segmentation (HP)
100% 1637
26.5 IPS
 
Image Segmentation (Q)
98% 3669
59.7 IPS
 
Pose Estimation (SP)
100% 3934
4.59 IPS
 
Pose Estimation (HP)
100% 3812
4.45 IPS
 
Pose Estimation (Q)
84% 10376
12.3 IPS
 
Object Detection (SP)
98% 1013
80.6 IPS
 
Object Detection (HP)
98% 1100
87.5 IPS
 
Object Detection (Q)
83% 2732
220.7 IPS
 
Face Detection (SP)
100% 2978
35.4 IPS
 
Face Detection (HP)
100% 2974
35.3 IPS
 
Face Detection (Q)
95% 6232
74.4 IPS
 
Depth Estimation (SP)
99% 3396
26.2 IPS
 
Depth Estimation (HP)
99% 2468
19.1 IPS
 
Depth Estimation (Q)
64% 4215
38.6 IPS
 
Style Transfer (SP)
89% 7932
10.3 IPS
 
Style Transfer (HP)
89% 6374
8.27 IPS
 
Style Transfer (Q)
98% 12424
16.0 IPS
 
Image Super-Resolution (SP)
100% 1424
52.6 IPS
 
Image Super-Resolution (HP)
100% 1488
54.9 IPS
 
Image Super-Resolution (Q)
97% 3239
120.0 IPS
 
Text Classification (SP)
100% 415
554.4 IPS
 
Text Classification (HP)
99% 422
562.6 IPS
 
Text Classification (Q)
88% 667
899.4 IPS
 
Machine Translation (SP)
100% 1061
18.3 IPS
 
Machine Translation (HP)
100% 900
15.5 IPS
 
Machine Translation (Q)
50% 251
8.90 IPS
 

Compare

Set Baseline Geekbench 6 now available