Deep Learning Accelerator Programming Interface (nvm_dlaSample)

The NvMedia nvm_dlaSample sample application demonstrates how to use the NvMedia Deep Learning Accelerator (DLA) APIs to perform deep learning inference operations. The sample uses the NVIDIA® SoC DLA hardware engine.

The nvm_dlaSample application has four testing modes:

  1. Runtime mode.
    • Single thread for running DLA.
    • Demonstrates how to create and initialize DLA instances.
    • Demonstrates how to run DLA with provided input data and network.
  2. SciSync mode.
    • Single thread for running DLA. Two supporting threads for synchronization (signaler and waiter).
    • Demonstrates how to create and initialize DLA instances.
    • Demonstrates how to run DLA with provided input data and network.
    • Demonstrates how to synchronize DLA task submission with a CPU signaler and CPU waiter.
  3. Multithreaded mode.
    • Multiple threads (4) for running DLA.
    • Demonstrates how to create and initialize DLA instances.
    • Demonstrates how to run DLA with provided input data and network.
  4. Ping mode.
    • Pings the specified instance if it exists.
Note: The tegrastats utility enables and reports on resource utilization. For addition information, refer to DLA in GitHub.