Hi Lunar, Thank you for reaching out with these questions - will do my best to answer them. Happy to see you are interested in this topic.
1 - we can look into finding a machine with a cuda supported GPU, but this will probably be easier if you have your own. 2 - personally I find all of the subproject areas interesting and would be useful to the project, so it is hard to assign priority. It would be beneficial as you say to propose a development sequence based on feasibility and impact. 3 - unit tests for individual GPU-assisted computation blocks will be sufficient. Hope this helps, and look forward to your proposal. Josh On Wed, Mar 12, 2025 at 3:14 AM Mohammad Mosiur Lunar <mmlu...@hotmail.com> wrote: > Hello Everyone, > I hope this email finds you well. My name is Lunar, and I am currently a > Ph.D. student at the University of Nebraska-Lincoln, specializing in > wireless communication experimentation and data analysis using deep > learning. Through my research, I have gained extensive experience with GNU > Radio and USRP devices, and I am particularly interested in leveraging GPU > acceleration for real-time signal processing. > The GPU-Accelerated Signal Processing Blocks project strongly aligns with > my research interests, and I am eager to contribute. Before submitting my > detailed proposal, I would appreciate some clarifications regarding the > project scope: > > 1. Hardware/Computing Resources: Since this project requires GPU > acceleration, will GNU Radio provide contributors with access to a > CUDA-supported GPU, either via a development machine or through a cloud > computing service? Or is it expected that contributors arrange this > themselves? > 2. Task Prioritization: The project outlines several subprojects. Is > there a predefined priority among them, or should the contributor propose a > development sequence based on feasibility and impact? > 3. Testing Scope: Will unit tests for individual GPU-assisted > computation blocks be sufficient, or is there an expectation to include > integration tests with USRP devices to validate end-to-end performance? > > Gaining clarity on these points will help me structure my proposal > effectively. I appreciate your time and look forward to your guidance. > Best regards, > Lunar > >
