User | Katyusha |
Upload Date | January 02 2025 03:27 AM |
Views | 8 |
AI Information | |
---|---|
Framework | TensorFlow Lite |
Backend | CPU |
Device | ZHAOXIN KaiXian KX-7000 |
System Information | |
---|---|
Operating System | AOSC OS (12.0.1) |
Model | Shanghai Zhaoxin Semiconductor Co., Ltd. ZXE CRB |
Motherboard | Shanghai Zhaoxin Semiconductor Co., Ltd. JZ7KMA01 |
CPU Information | |
---|---|
Name | ZHAOXIN KaiXian KX-7000 |
Topology | 1 Processor, 8 Cores |
Identifier | CentaurHauls Family 7 Model 11 Stepping 1 |
Base Frequency | 3.73 GHz |
Cluster 1 | 0 Cores |
Memory Information | |
---|---|
Size | 31.13 GB |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (SP)
|
100% |
615
114.4 IPS |
|
Image Classification (HP)
|
100% |
543
101.1 IPS |
|
Image Classification (Q)
|
99% |
437
81.5 IPS |
|
Image Segmentation (SP)
|
100% |
781
12.7 IPS |
|
Image Segmentation (HP)
|
100% |
782
12.7 IPS |
|
Image Segmentation (Q)
|
98% |
468
7.60 IPS |
|
Pose Estimation (SP)
|
100% |
868
1.01 IPS |
|
Pose Estimation (HP)
|
100% |
865
1.01 IPS |
|
Pose Estimation (Q)
|
96% |
1436
1.68 IPS |
|
Object Detection (SP)
|
100% |
527
41.8 IPS |
|
Object Detection (HP)
|
100% |
534
42.3 IPS |
|
Object Detection (Q)
|
85% |
423
34.1 IPS |
|
Face Detection (SP)
|
100% |
1251
14.9 IPS |
|
Face Detection (HP)
|
100% |
1255
14.9 IPS |
|
Face Detection (Q)
|
97% |
1033
12.3 IPS |
|
Depth Estimation (SP)
|
100% |
975
7.51 IPS |
|
Depth Estimation (HP)
|
99% |
979
7.55 IPS |
|
Depth Estimation (Q)
|
63% |
928
8.70 IPS |
|
Style Transfer (SP)
|
100% |
1333
1.71 IPS |
|
Style Transfer (HP)
|
100% |
1335
1.72 IPS |
|
Style Transfer (Q)
|
98% |
2539
3.27 IPS |
|
Image Super-Resolution (SP)
|
100% |
619
22.9 IPS |
|
Image Super-Resolution (HP)
|
100% |
617
22.8 IPS |
|
Image Super-Resolution (Q)
|
97% |
773
28.7 IPS |
|
Text Classification (SP)
|
100% |
672
897.3 IPS |
|
Text Classification (HP)
|
100% |
670
894.9 IPS |
|
Text Classification (Q)
|
92% |
289
388.1 IPS |
|
Machine Translation (SP)
|
100% |
841
14.5 IPS |
|
Machine Translation (HP)
|
100% |
876
15.1 IPS |
|
Machine Translation (Q)
|
58% |
265
6.22 IPS |