Geekbench AI Score

235
Single Precision Score
324
Half Precision Score
325
Quantized Score
Geekbench AI 1.2.0 for Android AArch64

Result Information

User kitsunechan_2525
Upload Date June 10 2025 03:47 AM
Views 11

System Information

AI Information
Framework TensorFlow Lite
Backend GPU
Device ARM MT6877V/TTZA
System Information
Operating System Android 14
Model Xiaomi 22101316C
Model IDXiaomi 22101316C
Motherboard ruby
CPU Information
Name ARM MT6877V/TTZA
Topology 1 Processor, 8 Cores
Identifier ARM implementer 65 architecture 8 variant 1 part 3393 revision 0
Base Frequency 2.00 GHz
Cluster 16 Cores @ 2.00 GHz
Cluster 22 Cores @ 2.60 GHz
Memory Information
Size 7.37 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 167
31.1 IPS
 
Image Classification (HP)
100% 287
53.3 IPS
 
Image Classification (Q)
99% 228
42.6 IPS
 
Image Segmentation (SP)
100% 229
3.71 IPS
 
Image Segmentation (HP)
100% 402
6.52 IPS
 
Image Segmentation (Q)
98% 382
6.21 IPS
 
Pose Estimation (SP)
100% 161
0.19 IPS
 
Pose Estimation (HP)
100% 189
0.22 IPS
 
Pose Estimation (Q)
96% 188
0.22 IPS
 
Object Detection (SP)
100% 132
10.4 IPS
 
Object Detection (HP)
100% 206
16.4 IPS
 
Object Detection (Q)
83% 180
14.6 IPS
 
Face Detection (SP)
100% 513
6.10 IPS
 
Face Detection (HP)
100% 834
9.91 IPS
 
Face Detection (Q)
97% 705
8.41 IPS
 
Depth Estimation (SP)
100% 308
2.37 IPS
 
Depth Estimation (HP)
99% 397
3.06 IPS
 
Depth Estimation (Q)
62% 321
3.04 IPS
 
Style Transfer (SP)
100% 622
0.80 IPS
 
Style Transfer (HP)
100% 875
1.12 IPS
 
Style Transfer (Q)
98% 870
1.12 IPS
 
Image Super-Resolution (SP)
100% 157
5.79 IPS
 
Image Super-Resolution (HP)
100% 238
8.77 IPS
 
Image Super-Resolution (Q)
97% 236
8.75 IPS
 
Text Classification (SP)
100% 162
216.4 IPS
 
Text Classification (HP)
100% 163
217.8 IPS
 
Text Classification (Q)
91% 361
485.4 IPS
 
Machine Translation (SP)
100% 258
4.44 IPS
 
Machine Translation (HP)
100% 255
4.39 IPS
 
Machine Translation (Q)
57% 266
6.51 IPS
 

Compare

Set Baseline Geekbench 6 now available