Profile Job Results
Jobs
j0pxjyj8g
Results Ready
Name
controlnet_quantized_UNet_Quantized
Target Device
- QCS8550 (Proxy)
- Android 12
- Qualcomm® QCS8550
Creator
ai-hub-support@qti.qualcomm.com
Target Model
Input Specs
input_1: uint16[1, 64, 64, 4]input_2: uint16[1, 1280]input_3: uint16[1, 77, 768]controlnet_downblock1: uint16[1, 64, 64, 320]controlnet_downblock2: uint16[1, 64, 64, 320]controlnet_downblock3: uint16[1, 64, 64, 320]controlnet_downblock4: uint16[1, 32, 32, 320]controlnet_downblock5: uint16[1, 32, 32, 640]controlnet_downblock6: uint16[1, 32, 32, 640]controlnet_downblock7: uint16[1, 16, 16, 640]controlnet_downblock8: uint16[1, 16, 16, 1280]controlnet_downblock9: uint16[1, 16, 16, 1280]controlnet_downblock10: uint16[1, 8, 8, 1280]controlnet_downblock11: uint16[1, 8, 8, 1280]controlnet_downblock12: uint16[1, 8, 8, 1280]controlnet_midblock: uint16[1, 8, 8, 1280]Completion Time
10/3/2024, 11:30:22 PM
Versions
- QNN: v2.26.0.240827110523_99241
- QNN Backend API: 5.26.0
- QNN Core API: 2.19.0
- Android: 13 (TP1A.220624.014)
- AI Hub: aihub-2024.10.01.0
Estimated Inference Time
260 ms
Estimated Peak Memory Usage
14 ‑ 15 MB
Compute Units
NPU
5433
| Stage | Time | Memory |
|---|---|---|
First App Load | 1.00 s | 2‑3 MB |
Subsequent App Load | 550 ms | 1‑3 MB |
Inference | 260 ms | 14‑15 MB |
| QNN | Value |
|---|---|
| context_options.htp_options.performance_mode | BURST |
| default_graph_options.htp_options.precision | FLOAT16 |
Sign up to run this model on a hosted Qualcomm® device!
Run on device