User | Neredera |
Upload Date | February 20 2025 11:34 PM |
Views | 1 |
AI Information | |
---|---|
Framework | TensorFlow Lite |
Backend | CPU |
Device | AMD EPYC 9534 |
System Information | |
---|---|
Operating System | Debian GNU/Linux 12 (bookworm) |
Model | Supermicro Super Server |
Motherboard | Supermicro H13SSL-N |
CPU Information | |
---|---|
Name | AMD EPYC 9534 |
Topology | 1 Processor, 64 Cores |
Identifier | AuthenticAMD Family 25 Model 17 Stepping 1 |
Base Frequency | 3.72 GHz |
Cluster 1 | 0 Cores |
Memory Information | |
---|---|
Size | 31.05 GB |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (SP)
|
100% |
484
90.0 IPS |
|
Image Classification (HP)
|
100% |
521
96.8 IPS |
|
Image Classification (Q)
|
99% |
994
185.4 IPS |
|
Image Segmentation (SP)
|
100% |
957
15.5 IPS |
|
Image Segmentation (HP)
|
100% |
974
15.8 IPS |
|
Image Segmentation (Q)
|
98% |
1113
18.1 IPS |
|
Pose Estimation (SP)
|
100% |
1429
1.67 IPS |
|
Pose Estimation (HP)
|
100% |
1408
1.64 IPS |
|
Pose Estimation (Q)
|
96% |
3327
3.90 IPS |
|
Object Detection (SP)
|
100% |
488
38.7 IPS |
|
Object Detection (HP)
|
100% |
531
42.1 IPS |
|
Object Detection (Q)
|
85% |
1192
95.9 IPS |
|
Face Detection (SP)
|
100% |
775
9.21 IPS |
|
Face Detection (HP)
|
100% |
791
9.40 IPS |
|
Face Detection (Q)
|
97% |
1730
20.6 IPS |
|
Depth Estimation (SP)
|
100% |
773
5.96 IPS |
|
Depth Estimation (HP)
|
99% |
781
6.02 IPS |
|
Depth Estimation (Q)
|
63% |
1302
12.2 IPS |
|
Style Transfer (SP)
|
100% |
1550
1.99 IPS |
|
Style Transfer (HP)
|
100% |
1564
2.01 IPS |
|
Style Transfer (Q)
|
98% |
3248
4.19 IPS |
|
Image Super-Resolution (SP)
|
100% |
348
12.8 IPS |
|
Image Super-Resolution (HP)
|
100% |
383
14.1 IPS |
|
Image Super-Resolution (Q)
|
97% |
915
33.9 IPS |
|
Text Classification (SP)
|
100% |
434
579.8 IPS |
|
Text Classification (HP)
|
100% |
424
565.4 IPS |
|
Text Classification (Q)
|
92% |
310
416.1 IPS |
|
Machine Translation (SP)
|
100% |
906
15.6 IPS |
|
Machine Translation (HP)
|
100% |
870
15.0 IPS |
|
Machine Translation (Q)
|
58% |
492
11.6 IPS |