QNN LPAI Backend Glossary¶
- LPAI¶
Low Power AI backend optimized for low-area, low-power applications such as always-on voice and camera use cases.
- QNN_LPAI_API_VERSION_MAJOR¶
Major version of the QNN LPAI backend API. Current: .
- QNN_LPAI_API_VERSION_MINOR¶
Minor version of the QNN LPAI backend API. Current: .
- QNN_LPAI_API_VERSION_PATCH¶
Patch version of the QNN LPAI backend API. Current: .
- QNN_SDK_ROOT¶
Environment variable pointing to the root directory of the QNN SDK installation.
- LD_LIBRARY_PATH¶
Linux environment variable specifying paths to shared libraries required by QNN tools.
- target_env¶
Specifies the target environment for model execution. Options: arm, adsp, x86.
- enable_hw_ver¶
Specifies the hardware version of the LPAI backend. Options: v5, v5_1, v6.
- fps¶
Frames per second setting for model execution. Default: 1.
- ftrt_ratio¶
Frame-to-real-time ratio. Default: 10.
- client_type¶
Type of workload. Options: real_time, non_real_time.
- affinity¶
Core affinity policy. Options: soft, hard.
- core_selection¶
Specifies the core number for execution. Default: 0.
- profiling_level¶
Level of profiling detail. Options: basic, detailed.
- is_persistent_binary¶
Indicates whether the context binary must persist until QnnContext_free is called.
- QnnContext_createFromBinary¶
API used to create a QNN context from a serialized binary.
- QnnGraph_finalize¶
API used to finalize a graph before execution.
- QnnGraph_execute¶
API used to execute a finalized graph.
- QnnContext_free¶
API used to release a QNN context.
- QnnLpaiMem_MemType¶
Enum defining memory types: DDR, LLC, TCM, UNDEFINED.
- QnnLpaiGraph_ClientPerfType¶
Enum defining client performance types: REAL_TIME, NON_REAL_TIME.
- QnnLpaiGraph_CoreAffinityType¶
Enum defining core affinity types: SOFT, HARD, UNDEFINED.
- Scratch Memory¶
Memory used for intermediate results that can be overwritten during execution.
- Persistent Memory¶
Memory used for intermediate results that must persist across operations.
- Backend Extension¶
JSON configuration enabling custom options for LPAI backend tools.
- qnn-net-run¶
Tool used to execute QNN models on supported platforms.
- qnn-context-binary-generator¶
Tool used to generate offline context binaries for QNN models.
- qnn-profile-viewer¶
Tool used to visualize the profiling data.
- QNN_CONVERTOR¶
Tool responsible for converting and quantizing models to QNN format.