RidgeRun Services Development Workflow

From RidgeRun Developer Wiki


Follow Us On Twitter LinkedIn Email Share this page



NVIDIA partner logo






RidgeRun Services Development Workflow: PVA

RidgeRun has developed a clear and effective development workflow to deliver successful PVA-based solutions:

1. Customer Contact & Requirements Gathering:

  • The Statement of Work (SoW) is received.
  • Optionally includes the algorithm to be ported (either in Python, CPU, or GPU), input/output format details, and performance expectations.

2. Feasibility Evaluation:

  • RidgeRun performs a technical analysis to determine whether the algorithm fits within PVA constraints (e.g., memory, data flow, supported operators).
  • If needed, RidgeRun proposes algorithm adaptations to achieve compatibility.

3. Implementation Phase:

  • A working version is implemented using a shared library in C/C++ or RidgeRun’s GStreamer VPI Plugin.
  • The algorithm is validated against functional and performance requirements.

4. Optimization Loop:

  • If performance improvements are required, RidgeRun iteratively applies optimization techniques.
  • Optimization continues until diminishing returns are observed or target performance is reached.

5. Delivery and Support:

  • RidgeRun provides the PVA configuration files in binary form, ensuring compatibility with the customer's deployment environment.
  • Optional post-delivery support includes integration assistance and further tuning.

For more information about our services, you can visit our website.




Retrieved from "https://developer.ridgerun.com/wiki/index.php?title=RidgeRun_NVIDIA_PVA_Development/Introduction/Development_Workflow&oldid=66208"