Geekbench AI Score

578
Single Precision Score
653
Half Precision Score
679
Quantized Score
Geekbench AI 1.2.0 for Android AArch64

Result Information

Upload Date December 28 2024 10:28 AM
Views 2

System Information

AI Information
Framework TensorFlow Lite
Backend GPU
Device ARM ARMv8
System Information
Operating System Android 15
Model OPPO CPH2519
Model IDOPPO CPH2519
Motherboard k6985v1_64
CPU Information
Name ARM ARMv8
Topology 1 Processor, 8 Cores
Identifier ARM implementer 65 architecture 8 variant 1 part 3406 revision 0
Base Frequency 1.80 GHz
Cluster 14 Cores @ 1.80 GHz
Cluster 23 Cores @ 2.85 GHz
Cluster 31 Core @ 3.05 GHz
Memory Information
Size 11.10 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 285
53.0 IPS
 
Image Classification (HP)
100% 295
54.8 IPS
 
Image Classification (Q)
99% 351
65.4 IPS
 
Image Segmentation (SP)
100% 786
12.7 IPS
 
Image Segmentation (HP)
100% 879
14.2 IPS
 
Image Segmentation (Q)
98% 818
13.3 IPS
 
Pose Estimation (SP)
100% 534
0.62 IPS
 
Pose Estimation (HP)
100% 735
0.86 IPS
 
Pose Estimation (Q)
96% 729
0.85 IPS
 
Object Detection (SP)
100% 264
21.0 IPS
 
Object Detection (HP)
100% 331
26.2 IPS
 
Object Detection (Q)
83% 389
31.5 IPS
 
Face Detection (SP)
100% 1377
16.4 IPS
 
Face Detection (HP)
100% 1389
16.5 IPS
 
Face Detection (Q)
97% 1303
15.5 IPS
 
Depth Estimation (SP)
100% 920
7.09 IPS
 
Depth Estimation (HP)
99% 1095
8.43 IPS
 
Depth Estimation (Q)
62% 895
8.47 IPS
 
Style Transfer (SP)
100% 2639
3.39 IPS
 
Style Transfer (HP)
100% 3329
4.28 IPS
 
Style Transfer (Q)
98% 3299
4.25 IPS
 
Image Super-Resolution (SP)
100% 444
16.4 IPS
 
Image Super-Resolution (HP)
100% 539
19.9 IPS
 
Image Super-Resolution (Q)
97% 722
26.8 IPS
 
Text Classification (SP)
100% 251
334.6 IPS
 
Text Classification (HP)
100% 238
317.1 IPS
 
Text Classification (Q)
91% 352
472.7 IPS
 
Machine Translation (SP)
100% 357
6.15 IPS
 
Machine Translation (HP)
100% 349
6.00 IPS
 
Machine Translation (Q)
57% 262
6.39 IPS
 

Compare

Set Baseline Geekbench 6 now available