Save the Date: January’s NRP Engagement Webinar on Running a 380PFLOP32s GPU Burst for Multi-Messenger Astrophysics with IceCube Across all Available GPUs in the Cloud
We are excited to kick off the New Year with the second National Research Platform (NPR) Engagement meeting series on Monday, January 27 at 1 p.m. ET.
The webinar will feature Frank Würthwein, executive director of the Open Science Grid and physics professor at UC San Diego, and Igor Sfiligoi (pictured on the right), lead scientific software developer and researcher at UCSD/SDSC.
With the help of an EAGER grant from the National Science Foundation to explore the feasibility of running a large spike of GPU-based computing in the Cloud, Frank and Igor recently succeeded in getting over 50k GPUs from the three largest Cloud providers, integrating them in a single HTCondor pool, and running production IceCube simulations on them.
Their presentation on January 27 will report on a few hour GPU burst across Amazon Web Services, Microsoft Azure, and Google Cloud Platform that harvested all available for sale GPUs across the three cloud providers the weekend before SC19, reaching over 51k GPUs total and 380 PFLOP32s.
They will also report on the scale and science performance achieved across all the various GPU types, as well as the science motivation to do so. Additional details about the webinar can be found on the NRP Wiki page.
I hope you and your colleagues will consider joining us for this exciting presentation on January 27.