Upload Date | January 17 2025 03:11 AM |
Views | 4 |
System Information | |
---|---|
Operating System | Gentoo Linux |
Model | Supermicro Super Server |
Motherboard | Supermicro X10SRi-F |
CPU Information | |
---|---|
Name | Intel Xeon E5-1650 v4 |
Topology | 1 Processor, 6 Cores |
Identifier | GenuineIntel Family 6 Model 79 Stepping 1 |
Base Frequency | 3.60 GHz |
Cluster 1 | 0 Cores |
L1 Instruction Cache | 32.0 KB x 3 |
L1 Data Cache | 32.0 KB x 3 |
L2 Cache | 256 KB x 3 |
L3 Cache | 15.0 MB x 1 |
Memory Information | |
---|---|
Size | 251.58 GB |
Inference Information | |
---|---|
Framework | TensorFlow Lite |
Backend | CPU |
Device | Default |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (F32)
|
100% |
108
20.2 IPS |
|
Image Classification (F16)
|
100% |
85
15.9 IPS |
|
Image Classification (I8)
|
99% |
99
18.5 IPS |
|
Image Segmentation (F32)
|
100% |
419
7.00 IPS |
|
Image Segmentation (F16)
|
100% |
404
6.75 IPS |
|
Image Segmentation (I8)
|
98% |
280
4.68 IPS |
|
Pose Estimation (F32)
|
100% |
1172
1.42 IPS |
|
Pose Estimation (F16)
|
100% |
1169
1.42 IPS |
|
Pose Estimation (I8)
|
100% |
1002
1.21 IPS |
|
Object Detection (F32)
|
100% |
177
13.2 IPS |
|
Object Detection (F16)
|
100% |
171
12.8 IPS |
|
Object Detection (I8)
|
65% |
158
11.8 IPS |
|
Face Detection (F32)
|
100% |
502
5.97 IPS |
|
Face Detection (F16)
|
100% |
504
5.99 IPS |
|
Face Detection (I8)
|
87% |
458
5.45 IPS |
|
Depth Estimation (F32)
|
100% |
542
4.20 IPS |
|
Depth Estimation (F16)
|
100% |
555
4.31 IPS |
|
Depth Estimation (I8)
|
95% |
484
3.75 IPS |
|
Style Transfer (F32)
|
100% |
1725
2.27 IPS |
|
Style Transfer (F16)
|
100% |
1705
2.24 IPS |
|
Style Transfer (I8)
|
98% |
2075
2.73 IPS |
|
Image Super-Resolution (F32)
|
100% |
312
11.2 IPS |
|
Image Super-Resolution (F16)
|
100% |
331
11.8 IPS |
|
Image Super-Resolution (I8)
|
98% |
299
10.7 IPS |
|
Text Classification (F32)
|
100% |
24
34.9 IPS |
|
Text Classification (F16)
|
100% |
20
28.1 IPS |
|
Text Classification (I8)
|
92% |
23
32.8 IPS |
|
Machine Translation (F32)
|
100% |
49
0.90 IPS |
|
Machine Translation (F16)
|
100% |
51
0.94 IPS |
|
Machine Translation (I8)
|
62% |
52
0.95 IPS |