User | LinFengJun |
Upload Date | August 18 2024 04:24 AM |
Views | 2 |
System Information | |
---|---|
Operating System | Android 12 |
Model | Xiaomi Redmi 9 |
Model ID | Xiaomi M2004J19C |
Motherboard | lancelot |
CPU Information | |
---|---|
Name | ARM MT6769T |
Topology | 1 Processor, 8 Cores |
Identifier | ARM implementer 65 architecture 8 variant 3 part 3338 revision 1 |
Base Frequency | 1.80 GHz |
Cluster 1 | 6 Cores @ 1.80 GHz |
Cluster 2 | 2 Cores @ 2.00 GHz |
Memory Information | |
---|---|
Size | 3.61 GB |
Inference Information | |
---|---|
Framework | TensorFlow Lite |
Backend | GPU |
Device | Default |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (F32)
|
100% |
70
13.1 IPS |
|
Image Classification (F16)
|
100% |
115
21.5 IPS |
|
Image Classification (I8)
|
97% |
108
20.1 IPS |
|
Image Segmentation (F32)
|
100% |
99
1.66 IPS |
|
Image Segmentation (F16)
|
100% |
181
3.02 IPS |
|
Image Segmentation (I8)
|
98% |
171
2.86 IPS |
|
Pose Estimation (F32)
|
100% |
54
0.07 IPS |
|
Pose Estimation (F16)
|
100% |
74
0.09 IPS |
|
Pose Estimation (I8)
|
100% |
74
0.09 IPS |
|
Object Detection (F32)
|
100% |
54
4.02 IPS |
|
Object Detection (F16)
|
100% |
86
6.39 IPS |
|
Object Detection (I8)
|
62% |
83
6.20 IPS |
|
Face Detection (F32)
|
100% |
222
2.64 IPS |
|
Face Detection (F16)
|
98% |
342
4.07 IPS |
|
Face Detection (I8)
|
86% |
325
3.86 IPS |
|
Depth Estimation (F32)
|
100% |
120
0.93 IPS |
|
Depth Estimation (F16)
|
100% |
164
1.27 IPS |
|
Depth Estimation (I8)
|
95% |
163
1.27 IPS |
|
Style Transfer (F32)
|
100% |
230
0.30 IPS |
|
Style Transfer (F16)
|
100% |
348
0.46 IPS |
|
Style Transfer (I8)
|
98% |
346
0.46 IPS |
|
Image Super-Resolution (F32)
|
100% |
67
2.41 IPS |
|
Image Super-Resolution (F16)
|
100% |
102
3.66 IPS |
|
Image Super-Resolution (I8)
|
98% |
102
3.63 IPS |
|
Text Classification (F32)
|
100% |
63
91.0 IPS |
|
Text Classification (F16)
|
100% |
64
91.4 IPS |
|
Text Classification (I8)
|
92% |
96
138.0 IPS |
|
Machine Translation (F32)
|
100% |
121
2.24 IPS |
|
Machine Translation (F16)
|
100% |
121
2.22 IPS |
|
Machine Translation (I8)
|
64% |
131
2.41 IPS |