Geekbench ML Score

1067
TensorFlow Lite GPU Inference Score
Geekbench ML 0.6.0 for Android AArch64

Result Information

Upload Date October 18 2025 03:14 PM
Views 6

System Information

System Information
Operating System Android 15
Model vivo V2405A
Model IDvivo V2405A
Motherboard k6991v1_64
CPU Information
Name ARM ARMv8
Topology 1 Processor, 8 Cores
Identifier ARM implementer 65 architecture 8 variant 0 part 3461 revision 1
Base Frequency 2.40 GHz
Cluster 14 Cores @ 2.40 GHz
Cluster 23 Cores @ 3.30 GHz
Cluster 31 Core @ 3.63 GHz
Memory Information
Size 11.08 GB
Inference Information
Framework TensorFlow Lite
Backend GPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 773
144.6 IPS
 
Image Classification (F16)
100% 530
99.3 IPS
 
Image Classification (I8)
97% 469
87.8 IPS
 
Image Segmentation (F32)
100% 1127
18.8 IPS
 
Image Segmentation (F16)
100% 1349
22.5 IPS
 
Image Segmentation (I8)
98% 1345
22.5 IPS
 
Pose Estimation (F32)
100% 951
1.15 IPS
 
Pose Estimation (F16)
100% 1503
1.82 IPS
 
Pose Estimation (I8)
100% 1207
1.46 IPS
 
Object Detection (F32)
100% 429
32.0 IPS
 
Object Detection (F16)
100% 494
36.9 IPS
 
Object Detection (I8)
62% 487
36.3 IPS
 
Face Detection (F32)
100% 2053
24.4 IPS
 
Face Detection (F16)
98% 2325
27.6 IPS
 
Face Detection (I8)
86% 2153
25.6 IPS
 
Depth Estimation (F32)
100% 1280
9.92 IPS
 
Depth Estimation (F16)
100% 1594
12.4 IPS
 
Depth Estimation (I8)
95% 1566
12.1 IPS
 
Style Transfer (F32)
100% 3702
4.87 IPS
 
Style Transfer (F16)
100% 4922
6.47 IPS
 
Style Transfer (I8)
98% 4867
6.40 IPS
 
Image Super-Resolution (F32)
100% 713
25.5 IPS
 
Image Super-Resolution (F16)
100% 843
30.1 IPS
 
Image Super-Resolution (I8)
98% 919
32.8 IPS
 
Text Classification (F32)
100% 493
709.2 IPS
 
Text Classification (F16)
100% 489
702.5 IPS
 
Text Classification (I8)
92% 679
975.5 IPS
 
Machine Translation (F32)
100% 731
13.5 IPS
 
Machine Translation (F16)
100% 791
14.6 IPS
 
Machine Translation (I8)
64% 671
12.3 IPS
 

Compare

Set Baseline Geekbench 6 now available