User | _Aidan |
Upload Date | September 02 2024 06:27 AM |
Views | 2 |
System Information | |
---|---|
Operating System | Microsoft Windows 11 Pro (64-bit) |
Model | Gigabyte Technology Co., Ltd. TRX50 AERO D |
Motherboard | Gigabyte Technology Co., Ltd. TRX50 AERO D |
Power Plan | Balanced |
CPU Information | |
---|---|
Name | AMD Ryzen Threadripper 7960X 24-Cores |
Topology | 1 Processor, 24 Cores, 48 Threads |
Identifier | AuthenticAMD Family 25 Model 24 Stepping 1 |
Base Frequency | 4.20 GHz |
Cluster 1 | 24 Cores |
L1 Instruction Cache | 32.0 KB x 24 |
L1 Data Cache | 32.0 KB x 24 |
L2 Cache | 1.00 MB x 24 |
L3 Cache | 32.0 MB x 4 |
Memory Information | |
---|---|
Size | 128.00 GB |
Inference Information | |
---|---|
Framework | ONNX |
Backend | DirectML |
Device | NVIDIA GeForce RTX 4090 |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (F32)
|
100% |
10581
1.98 KIPS |
|
Image Classification (F16)
|
100% |
10636
1.99 KIPS |
|
Image Classification (I8)
|
100% |
8985
1.68 KIPS |
|
Image Segmentation (F32)
|
100% |
17876
298.5 IPS |
|
Image Segmentation (F16)
|
100% |
17807
297.4 IPS |
|
Image Segmentation (I8)
|
98% |
16998
283.9 IPS |
|
Pose Estimation (F32)
|
100% |
251123
304.1 IPS |
|
Pose Estimation (F16)
|
100% |
249228
301.8 IPS |
|
Pose Estimation (I8)
|
100% |
217807
263.8 IPS |
|
Object Detection (F32)
|
100% |
7839
585.2 IPS |
|
Object Detection (F16)
|
100% |
7568
565.0 IPS |
|
Object Detection (I8)
|
54% |
7283
543.7 IPS |
|
Face Detection (F32)
|
100% |
38284
455.2 IPS |
|
Face Detection (F16)
|
100% |
38546
458.3 IPS |
|
Face Detection (I8)
|
88% |
32544
387.0 IPS |
|
Depth Estimation (F32)
|
100% |
59472
461.2 IPS |
|
Depth Estimation (F16)
|
100% |
59673
462.8 IPS |
|
Depth Estimation (I8)
|
95% |
47623
369.3 IPS |
|
Style Transfer (F32)
|
100% |
365729
481.1 IPS |
|
Style Transfer (F16)
|
100% |
377635
496.7 IPS |
|
Style Transfer (I8)
|
98% |
333253
438.4 IPS |
|
Image Super-Resolution (F32)
|
100% |
43945
1.57 KIPS |
|
Image Super-Resolution (F16)
|
100% |
43254
1.54 KIPS |
|
Image Super-Resolution (I8)
|
99% |
34720
1.24 KIPS |
|
Text Classification (F32)
|
100% |
2833
4.07 KIPS |
|
Text Classification (F16)
|
100% |
2866
4.12 KIPS |
|
Text Classification (I8)
|
98% |
1724
2.48 KIPS |
|
Machine Translation (F32)
|
100% |
4579
84.3 IPS |
|
Machine Translation (F16)
|
100% |
4654
85.7 IPS |
|
Machine Translation (I8)
|
70% |
2470
45.5 IPS |