Profile Job Results
Jobs
jp1ex9d8g
Results Ready
Name
sam2_w8a8_SAM2Decoder
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
Target Model
Input Specs
image_embeddings: uint8[1, 64, 64, 256]high_res_features1: uint8[1, 256, 256, 32]high_res_features2: uint8[1, 128, 128, 64]sparse_embedding: uint8[1, 3, 256]Completion Time
1/10/2026, 7:25:05 PM
Options
--qairt_version latestVersions
- TensorFlow Lite: 2.17.0
- QAIRT: v2.41.0.251128145156_191518
- QNN TfLite Delegate: v2.41.0.251128145156_191518
- Qualcomm Linux: 1.6-ver.1.2
- AI Hub: aihub-2026.01.05.0
Estimated Inference Time
38.8 ms
Estimated Peak Memory Usage
14 ‑ 60 MB
Compute Units
NPU
740
CPU
70
GPU
20
| Stage | Time | Memory |
|---|---|---|
First App Load | 10.7 s | 278‑279 MB |
Subsequent App Load | 409 ms | 24‑25 MB |
Inference | 38.8 ms | 14‑60 MB |
| TensorFlow Lite | Value |
|---|---|
| number_of_threads | 4 |
| QNN Delegate | Value |
|---|---|
| backend_type | kHtpBackend |
| log_level | kLogLevelWarn |
| htp_options.performance_mode | kHtpBurst |
| htp_options.precision | kHtpFp16 |
| htp_options.optimization_strategy | kHtpOptimizeForInference |
| htp_options.useConvHmx | true |
| 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