Bob Johnson in the Environmental Science division at ANL has organized a workshop on the convergence of HPC and Geospatial Analysis. The meeting has been a good opportunity to learn one another’s language and concerns, and to consider how current and future HPC platforms might be employed to address computationally intensive components of geospatial workflows. My presentation, a short intro to HPC storage and analysis, is available here.
The Lustre User Group organizers kindly invited me to participate on a panel with Al Geist, Lee Ward, and Terri Quinn regarding future HPC systems and the role of parallel file systems. Great discussion! Slides should be available here soon.
For the second year, Argonne is organizing a two-week program on extreme scale computing. This year the program will be held on August 3-15, 2014 in the Chicago area. This is a unique program that brings together experts on tools and techniques in extreme-scale computing to give students an understanding of how things really work on big systems and for complex science codes. Rob Latham and I have organized one day of discussion around data, with our colleagues Avery Ching of Facebook, Quincey Koziol of the HDF Group, and Steve Tuecke of the University of Chicago.
The program is free, and domestic airfare, meals, and lodging are provided. Prospective students should apply here. Deadline is March 31st!
Hank Hoffmann and I are organizing the 3rd Annual Greater Chicago Area System Research Workshop (GCASR 2014) at the University of Chicago on May 19th, 2014. The workshop is a great place for researchers from all around the Chicago area, to meet and share their experiences. The format includes presentations and poster sessions, and submissions for both are being accepted until April 11th.
If you’re in the general vicinity and are working in systems research, we hope you will consider attending and submitting a presentation or poster on your work!
Today I had the opportunity to speak at an Office of Science and Technology Policy event on “Data to Knowledge to Action”. Stan Ahalt (UNC) ran a great breakout on R&D activities for Big Data in Science, and there was an especially good discussion of some of the technical/social/policy issues surrounding big data in a variety of contexts. More info here.
Ioan Raicu was kind enough to invite me to lecture in his data intensive computing course earlier this week. I gave a high-level view of some of the issues in HPC I/O systems targeting computational science. Questions were good, which is always a good sign. Slides can be found here.
While I’ll unfortunately be tied up much of the day in our I/O tutorial, we received some great papers this year, and I encourage you to check out the workshop on Monday Nov. 18th.
I participated in a review back in April with the goal of better understanding the needs of scientists working on DOE-sponsored experiments and facilities. Arie Shoshani and Bruce Hendrickson put together a nice report summarizing the review.
Rob Latham, Katie Antypas, Brent Welch, and I will be presenting our “Parallel I/O in Practice” tutorial at SC again this year, on Monday Nov. 18th. We always have some great discussions and are looking forward to the event!
IASDS is being held in conjunction with IEEE Cluster in Indianapolis, IN this year. It’s the fifth year for the workshop, and Dries and Doug Thain are organizing. The format usually leads to some good discussions.