Upload Date | August 16 2024 03:07 PM |
Views | 2 |
System Information | |
---|---|
Operating System | Microsoft Windows 11 Pro (64-bit) |
Model | Gigabyte Technology Co., Ltd. TRX40 DESIGNARE |
Motherboard | Gigabyte Technology Co., Ltd. TRX40 DESIGNARE |
Power Plan | Power saver |
CPU Information | |
---|---|
Name | AMD Ryzen Threadripper 3960X |
Topology | 1 Processor, 24 Cores, 48 Threads |
Identifier | AuthenticAMD Family 23 Model 49 Stepping 0 |
Base Frequency | 3.79 GHz |
Cluster 1 | 24 Cores |
L1 Instruction Cache | 32.0 KB x 24 |
L1 Data Cache | 32.0 KB x 24 |
L2 Cache | 512 KB x 24 |
L3 Cache | 16.0 MB x 8 |
Memory Information | |
---|---|
Size | 224.00 GB |
Inference Information | |
---|---|
Framework | ONNX |
Backend | DirectML |
Device | NVIDIA GeForce RTX 3090 Ti |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (F32)
|
100% |
6923
1.30 KIPS |
|
Image Classification (F16)
|
100% |
6591
1.23 KIPS |
|
Image Classification (I8)
|
100% |
4768
892.3 IPS |
|
Image Segmentation (F32)
|
100% |
11676
195.0 IPS |
|
Image Segmentation (F16)
|
100% |
11009
183.9 IPS |
|
Image Segmentation (I8)
|
98% |
8941
149.3 IPS |
|
Pose Estimation (F32)
|
100% |
172097
208.4 IPS |
|
Pose Estimation (F16)
|
100% |
175391
212.4 IPS |
|
Pose Estimation (I8)
|
100% |
127309
154.2 IPS |
|
Object Detection (F32)
|
100% |
2768
206.7 IPS |
|
Object Detection (F16)
|
100% |
2695
201.2 IPS |
|
Object Detection (I8)
|
56% |
2454
183.2 IPS |
|
Face Detection (F32)
|
100% |
22882
272.1 IPS |
|
Face Detection (F16)
|
100% |
23521
279.7 IPS |
|
Face Detection (I8)
|
89% |
16468
195.8 IPS |
|
Depth Estimation (F32)
|
100% |
51227
397.3 IPS |
|
Depth Estimation (F16)
|
100% |
51419
398.8 IPS |
|
Depth Estimation (I8)
|
94% |
33232
257.7 IPS |
|
Style Transfer (F32)
|
100% |
240628
316.5 IPS |
|
Style Transfer (F16)
|
100% |
237982
313.0 IPS |
|
Style Transfer (I8)
|
98% |
174424
229.4 IPS |
|
Image Super-Resolution (F32)
|
100% |
28882
1.03 KIPS |
|
Image Super-Resolution (F16)
|
100% |
29266
1.05 KIPS |
|
Image Super-Resolution (I8)
|
99% |
18877
674.2 IPS |
|
Text Classification (F32)
|
100% |
1772
2.55 KIPS |
|
Text Classification (F16)
|
100% |
1761
2.53 KIPS |
|
Text Classification (I8)
|
98% |
1025
1.47 KIPS |
|
Machine Translation (F32)
|
100% |
3115
57.3 IPS |
|
Machine Translation (F16)
|
100% |
3020
55.6 IPS |
|
Machine Translation (I8)
|
70% |
1427
26.3 IPS |