Geekbench AI Score

1091
Single Precision Score
1007
Half Precision Score
1483
Quantized Score
Geekbench AI 1.4.0 for Android AArch64

Result Information

Upload Date August 21 2025 07:19 AM
Views 4

System Information

AI Information
Framework TensorFlow Lite
Backend CPU
Device Qualcomm SM8750
System Information
Operating System Android 15
Model HONOR MBH-N49
Model IDHONOR MBH-N49
Motherboard MBH
Governorwalt
CPU Information
Name Qualcomm SM8750
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 14.89 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 1564
290.9 IPS
 
Image Classification (HP)
100% 758
140.9 IPS
 
Image Classification (Q)
100% 1170
217.7 IPS
 
Image Segmentation (SP)
100% 923
15.0 IPS
 
Image Segmentation (HP)
100% 861
14.0 IPS
 
Image Segmentation (Q)
98% 1433
23.3 IPS
 
Pose Estimation (SP)
100% 1759
2.05 IPS
 
Pose Estimation (HP)
100% 1758
2.05 IPS
 
Pose Estimation (Q)
84% 3726
4.42 IPS
 
Object Detection (SP)
98% 619
49.3 IPS
 
Object Detection (HP)
98% 691
54.9 IPS
 
Object Detection (Q)
83% 1112
89.8 IPS
 
Face Detection (SP)
100% 1689
20.1 IPS
 
Face Detection (HP)
100% 2036
24.2 IPS
 
Face Detection (Q)
95% 2822
33.7 IPS
 
Depth Estimation (SP)
99% 1965
15.2 IPS
 
Depth Estimation (HP)
99% 1995
15.4 IPS
 
Depth Estimation (Q)
64% 2865
26.2 IPS
 
Style Transfer (SP)
89% 3713
4.82 IPS
 
Style Transfer (HP)
89% 3680
4.78 IPS
 
Style Transfer (Q)
98% 7213
9.30 IPS
 
Image Super-Resolution (SP)
100% 810
29.9 IPS
 
Image Super-Resolution (HP)
100% 828
30.6 IPS
 
Image Super-Resolution (Q)
97% 1706
63.2 IPS
 
Text Classification (SP)
100% 289
386.3 IPS
 
Text Classification (HP)
99% 232
309.3 IPS
 
Text Classification (Q)
88% 295
397.6 IPS
 
Machine Translation (SP)
100% 530
9.13 IPS
 
Machine Translation (HP)
100% 475
8.18 IPS
 
Machine Translation (Q)
50% 253
9.00 IPS
 

Compare

Set Baseline Geekbench 6 now available