My name is Chung-Yuan Liang. I am a Ph.D. candidate working with Prof. Venkatesh Merwade in Civil Engineering at Purdue University. Concurrently, I am also studying for a Master’s degree in Computer Engineering. My research focuses on processing geospatial data and developing deep learning models for hydraulic and hydrologic modeling. I am particularly interested in employing high-performance computing and machine learning to advance geospatial applications and research.
Last week, at the AWRA 2024 GWTC conference, I met Prof. Huidae Cho, who introduced me to the Google Summer of Code (GSoC) program. Inspired by our conversation, I am writing to express my interest in contributing to the OSGeo community through the GSoC project. I am particularly drawn to the project titled “Parallelization of Existing Tools,” which Prof. Cho is mentoring.
Upon reviewing the issues and repositories of GRASS GIS, I had some preliminary ideas on parallelizing tools. My project proposal outlines the concepts and details of these ideas. I am sharing my proposal and would greatly appreciate any feedback or suggestions.
I’m glad that you’re interested in this parallelization project. FYI, the deadline for your application is April 2, 18:00 UTC (2pm tomorrow your time in EDT). Please check this feature request at https://github.com/OSGeo/grass/issues/2644 for testing your skills and let me know if you have any questions.
For whatever it is worth, it always struck me that v.surf.bspline was
low hanging fruit for OpenMP parallelization, as the code already
splits the data up into a manageable number of overlapping
tiles/segments, then processes them serially before recombining them.
Hi Hamish, good to hear from you! Thanks for pointing that out, I added it to the GSoC ideas wiki, so in case it won’t happen this year, we can try later.
For whatever it is worth, it always struck me that v.surf.bspline was
low hanging fruit for OpenMP parallelization, as the code already
splits the data up into a manageable number of overlapping
tiles/segments, then processes them serially before recombining them.
I just submitted a pull request to address issue #2644. I spent some time familiarizing myself with the source code of Grass GIS. I hope I am on the right track. If you have any comments or suggestions, please feel free to let me know. Thanks a lot!