Profile Job Results
Jobs
j5wlx7m4p
Results Ready
Name
lama_dilated
Target Device
- QCS8550 (Proxy)
- Android 12
- Qualcomm® Dragonwing™ QCS8550
Creator
ai-hub-support@qti.qualcomm.com
Input Specs
image: float32[1, 3, 512, 512]mask: float32[1, 1, 512, 512]Completion Time
9/19/2026, 8:11:37 PM
Options
--qairt_version latestVersions
- ONNX Runtime: 1.27.1
- QAIRT: v2.50.0.260828221209
- Android: 14 (UP1A.231005.007)
- AI Hub: aihub-2026.09.11.0
Estimated Inference Time
77.8 ms
Estimated Peak Memory Usage
0 ‑ 94 MB
Compute Units
NPU
282
| Stage | Time | Memory |
|---|---|---|
First App Load | 14.8 s | 642‑644 MB |
Subsequent App Load | 505 ms | 46‑203 MB |
Inference | 77.8 ms | 0‑94 MB |
| ONNX Runtime | Value |
|---|---|
| execution_mode | SEQUENTIAL |
| intra_op_num_threads | 0 |
| inter_op_num_threads | 0 |
| enable_memory_pattern | false |
| enable_cpu_memory_arena | false |
| graph_optimization_level | ENABLE_ALL |
| QNN Execution Provider | Value |
|---|---|
| htp_performance_mode | "burst" |
| htp_graph_finalization_optimization_mode | "3" |
| enable_htp_fp16_precision | "1" |
| capture_network_visualizations | false |
| context_priority | "normal" |
| offload_graph_io_quantization | "1" |
Sign up to run this model on a hosted Qualcomm® device!
Run on device