Profile Job Results
Jobs
jp2rn1v4g
Results Ready
Name
whisper_medium_quantized_decoder
Target Device
- Dragonwing RB3 Gen 2 Vision Kit
- Qualcomm Linux 1.6
- Qualcomm® Dragonwing™ RB3 Gen 2 | QCS6490
Creator
ai-hub-support@qti.qualcomm.com
Input Specs
input_ids: int32[1, 1]k_cache_cross_0: uint8[16, 1, 64, 1500]k_cache_cross_1: uint8[16, 1, 64, 1500]k_cache_cross_10: uint8[16, 1, 64, 1500]k_cache_cross_11: uint8[16, 1, 64, 1500]k_cache_cross_12: uint8[16, 1, 64, 1500]k_cache_cross_13: uint8[16, 1, 64, 1500]k_cache_cross_14: uint8[16, 1, 64, 1500]k_cache_cross_15: uint8[16, 1, 64, 1500]k_cache_cross_16: uint8[16, 1, 64, 1500]k_cache_cross_17: uint8[16, 1, 64, 1500]k_cache_cross_18: uint8[16, 1, 64, 1500]k_cache_cross_19: uint8[16, 1, 64, 1500]k_cache_cross_2: uint8[16, 1, 64, 1500]k_cache_cross_20: uint8[16, 1, 64, 1500]k_cache_cross_21: uint8[16, 1, 64, 1500]k_cache_cross_22: uint8[16, 1, 64, 1500]k_cache_cross_23: uint8[16, 1, 64, 1500]k_cache_cross_3: uint8[16, 1, 64, 1500]k_cache_cross_4: uint8[16, 1, 64, 1500]k_cache_cross_5: uint8[16, 1, 64, 1500]k_cache_cross_6: uint8[16, 1, 64, 1500]k_cache_cross_7: uint8[16, 1, 64, 1500]k_cache_cross_8: uint8[16, 1, 64, 1500]k_cache_cross_9: uint8[16, 1, 64, 1500]k_cache_self_0_in: uint8[16, 1, 64, 199]k_cache_self_10_in: uint8[16, 1, 64, 199]k_cache_self_11_in: uint8[16, 1, 64, 199]k_cache_self_12_in: uint8[16, 1, 64, 199]k_cache_self_13_in: uint8[16, 1, 64, 199]k_cache_self_14_in: uint8[16, 1, 64, 199]k_cache_self_15_in: uint8[16, 1, 64, 199]k_cache_self_16_in: uint8[16, 1, 64, 199]k_cache_self_17_in: uint8[16, 1, 64, 199]k_cache_self_18_in: uint8[16, 1, 64, 199]k_cache_self_19_in: uint8[16, 1, 64, 199]k_cache_self_1_in: uint8[16, 1, 64, 199]k_cache_self_20_in: uint8[16, 1, 64, 199]k_cache_self_21_in: uint8[16, 1, 64, 199]k_cache_self_22_in: uint8[16, 1, 64, 199]k_cache_self_23_in: uint8[16, 1, 64, 199]k_cache_self_2_in: uint8[16, 1, 64, 199]k_cache_self_3_in: uint8[16, 1, 64, 199]k_cache_self_4_in: uint8[16, 1, 64, 199]k_cache_self_5_in: uint8[16, 1, 64, 199]k_cache_self_6_in: uint8[16, 1, 64, 199]k_cache_self_7_in: uint8[16, 1, 64, 199]k_cache_self_8_in: uint8[16, 1, 64, 199]k_cache_self_9_in: uint8[16, 1, 64, 199]position_ids: int32[1]v_cache_cross_0: uint8[16, 1, 1500, 64]v_cache_cross_1: uint8[16, 1, 1500, 64]v_cache_cross_10: uint8[16, 1, 1500, 64]v_cache_cross_11: uint8[16, 1, 1500, 64]v_cache_cross_12: uint8[16, 1, 1500, 64]v_cache_cross_13: uint8[16, 1, 1500, 64]v_cache_cross_14: uint8[16, 1, 1500, 64]v_cache_cross_15: uint8[16, 1, 1500, 64]v_cache_cross_16: uint8[16, 1, 1500, 64]v_cache_cross_17: uint8[16, 1, 1500, 64]v_cache_cross_18: uint8[16, 1, 1500, 64]v_cache_cross_19: uint8[16, 1, 1500, 64]v_cache_cross_2: uint8[16, 1, 1500, 64]v_cache_cross_20: uint8[16, 1, 1500, 64]v_cache_cross_21: uint8[16, 1, 1500, 64]v_cache_cross_22: uint8[16, 1, 1500, 64]v_cache_cross_23: uint8[16, 1, 1500, 64]v_cache_cross_3: uint8[16, 1, 1500, 64]v_cache_cross_4: uint8[16, 1, 1500, 64]v_cache_cross_5: uint8[16, 1, 1500, 64]v_cache_cross_6: uint8[16, 1, 1500, 64]v_cache_cross_7: uint8[16, 1, 1500, 64]v_cache_cross_8: uint8[16, 1, 1500, 64]v_cache_cross_9: uint8[16, 1, 1500, 64]v_cache_self_0_in: uint8[16, 1, 199, 64]v_cache_self_10_in: uint8[16, 1, 199, 64]v_cache_self_11_in: uint8[16, 1, 199, 64]v_cache_self_12_in: uint8[16, 1, 199, 64]v_cache_self_13_in: uint8[16, 1, 199, 64]v_cache_self_14_in: uint8[16, 1, 199, 64]v_cache_self_15_in: uint8[16, 1, 199, 64]v_cache_self_16_in: uint8[16, 1, 199, 64]v_cache_self_17_in: uint8[16, 1, 199, 64]v_cache_self_18_in: uint8[16, 1, 199, 64]v_cache_self_19_in: uint8[16, 1, 199, 64]v_cache_self_1_in: uint8[16, 1, 199, 64]v_cache_self_20_in: uint8[16, 1, 199, 64]v_cache_self_21_in: uint8[16, 1, 199, 64]v_cache_self_22_in: uint8[16, 1, 199, 64]v_cache_self_23_in: uint8[16, 1, 199, 64]v_cache_self_2_in: uint8[16, 1, 199, 64]v_cache_self_3_in: uint8[16, 1, 199, 64]v_cache_self_4_in: uint8[16, 1, 199, 64]v_cache_self_5_in: uint8[16, 1, 199, 64]v_cache_self_6_in: uint8[16, 1, 199, 64]v_cache_self_7_in: uint8[16, 1, 199, 64]v_cache_self_8_in: uint8[16, 1, 199, 64]v_cache_self_9_in: uint8[16, 1, 199, 64]attention_mask: uint16[1, 1, 1, 200]Completion Time
10/3/2026, 10:05:24 AM
Versions
- QAIRT: v2.50.0.260828221209
- QNN Backend API: 5.50.0
- QNN Core API: 2.39.0
- Qualcomm Linux: 1.6-ver.1.2
- AI Hub: aihub-2026.09.25.0
Estimated Inference Time
64.1 ms
Estimated Peak Memory Usage
77 ‑ 168 MB
Compute Units
NPU
7425
| Stage | Time | Memory |
|---|---|---|
First App Load | 483 ms | 4‑5 MB |
Subsequent App Load | 519 ms | 4‑5 MB |
Inference | 64.1 ms | 77‑168 MB |
| QNN | Value |
|---|---|
| context_options.htp_options.performance_mode | BURST |
| default_graph_options.htp_options.optimizations[0].type | FINALIZE_OPTIMIZATION_FLAG |
| default_graph_options.htp_options.optimizations[0].value | 3.0 |
| default_graph_options.htp_options.precision | FLOAT16 |
| default_graph_options.htp_options.vtcm_size | 0 |
Sign up to run this model on a hosted Qualcomm® device!
Run on device