Geekbench ML Score

275
TensorFlow Lite CPU Inference Score
Geekbench ML 0.6.0 for Android AArch64

Result Information

Upload Date January 21 2024 04:38 PM
Views 1

System Information

System Information
Operating System Android 13
Model Xiaomi 2201117TY
Model IDXiaomi 2201117TY
Motherboard spesn
Governorschedutil
CPU Information
Name ARM Qualcomm
Topology 1 Processor, 8 Cores
Identifier ARM implementer 65 architecture 8 variant 1 part 3337 revision 0
Base Frequency 1.90 GHz
Cluster 14 Cores @ 1.90 GHz
Cluster 24 Cores @ 2.40 GHz
Memory Information
Size 3.56 GB
Inference Information
Framework TensorFlow Lite
Backend CPU
Device Default

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 153
28.6 IPS
 
Image Classification (F16)
100% 160
30.0 IPS
 
Image Classification (I8)
97% 240
44.9 IPS
 
Image Segmentation (F32)
100% 205
3.42 IPS
 
Image Segmentation (F16)
100% 209
3.49 IPS
 
Image Segmentation (I8)
98% 234
3.91 IPS
 
Pose Estimation (F32)
100% 289
0.35 IPS
 
Pose Estimation (F16)
100% 284
0.34 IPS
 
Pose Estimation (I8)
100% 807
0.98 IPS
 
Object Detection (F32)
100% 164
12.2 IPS
 
Object Detection (F16)
100% 165
12.4 IPS
 
Object Detection (I8)
61% 272
20.3 IPS
 
Face Detection (F32)
100% 321
3.82 IPS
 
Face Detection (F16)
100% 326
3.87 IPS
 
Face Detection (I8)
86% 505
6.01 IPS
 
Depth Estimation (F32)
100% 302
2.34 IPS
 
Depth Estimation (F16)
100% 304
2.36 IPS
 
Depth Estimation (I8)
95% 560
4.35 IPS
 
Style Transfer (F32)
100% 284
0.37 IPS
 
Style Transfer (F16)
100% 426
0.56 IPS
 
Style Transfer (I8)
98% 1030
1.35 IPS
 
Image Super-Resolution (F32)
100% 191
6.82 IPS
 
Image Super-Resolution (F16)
100% 194
6.92 IPS
 
Image Super-Resolution (I8)
98% 381
13.6 IPS
 
Text Classification (F32)
100% 204
293.4 IPS
 
Text Classification (F16)
100% 201
288.9 IPS
 
Text Classification (I8)
92% 148
213.3 IPS
 
Machine Translation (F32)
100% 370
6.80 IPS
 
Machine Translation (F16)
100% 367
6.75 IPS
 
Machine Translation (I8)
64% 146
2.68 IPS
 

Compare

Set Baseline Geekbench 6 now available