Profile Job Results
Jobs
jpxlml6lp
Results Ready
Name
distil_whisper_encoder
Target Device
- QCS8550 (Proxy)
- Android 12
- Qualcomm® Dragonwing™ QCS8550
Creator
ai-hub-support@qti.qualcomm.com
Input Specs
input_features: float32[1, 80, 3000]Completion Time
9/19/2026, 8:50:18 PM
Options
--qairt_version latest --compute_unit gpu --max_profiler_iterations 10Versions
- Lite RT: 1.4.2
- Android: 14 (UP1A.231005.007)
- AI Hub: aihub-2026.09.11.0
Estimated Inference Time
652 ms
Estimated Peak Memory Usage
0 ‑ 318 MB
Compute Units
GPU
1358
CPU
8
| Stage | Time | Memory |
|---|---|---|
First App Load | 5.87 s | 1,001‑1,002 MB |
Subsequent App Load | 656 ms | 651‑727 MB |
Inference | 652 ms | 0‑318 MB |
| Lite RT | Value |
|---|---|
| number_of_threads | 4 |
| GPUv2 Delegate | Value |
|---|---|
| inference_preference | TFLITE_GPU_INFERENCE_PREFERENCE_SUSTAINED_SPEED |
| inference_priority1 | TFLITE_GPU_INFERENCE_PRIORITY_MIN_LATENCY |
| inference_priority2 | TFLITE_GPU_INFERENCE_PRIORITY_MIN_MEMORY_USAGE |
| inference_priority3 | TFLITE_GPU_INFERENCE_PRIORITY_MAX_PRECISION |
| XNNPACK Delegate | Value |
|---|
Sign up to run this model on a hosted Qualcomm® device!
Run on device