Daily Bulletin

New learning opportunities in GPU computing

November 16, 2021

CISL encourages current and future users of GPU computing resources to participate in training and code-development events like the NOAA AI Science Bootcamp on December 10. Bootcamp participants will learn the basics of GPU computing and artificial intelligence (AI) concepts, particularly as they relate to applying deep learning to addressing scientific problems. The deadline to apply is November 24.

Another GPU hackathon – hosted by NOAA and the Oak Ridge Leadership Computing Facility – is scheduled for February 28 to March 2, 2022, with one day set for introductions the prior week. Both AI and classic HPC-based modeling groups with codes at early- or late-stage development are invited to apply and submit their projects by December 22. Accepted groups will have the opportunity to port and/or optimize their code alongside GPU-computing mentors and experts from the larger community, including computational scientists and research software engineers at national labs and NVIDIA. Feel free to reach out to dhoward@ucar.edu if you have any questions about these opportunities or about the application process.

Users can also contribute to the development of new CISL-hosted GPU training by completing this survey by November 30. Results of the survey will directly influence development of a training series focused on the use of GPU technology in Earth system science research.