Pushing HPC Forward: Georgia Tech Heads to Supercomputing '16

Monday, November 14, 2016

New drugs, better weather forecasts, safer vehicles, and more sustainable development–this is just a tiny sampling of what can be achieved through high-performance computing (HPC).

However, despite gains in these and other domains in recent years, there is much more to be done before HPC can begin to reach its full potential.

That’s why a group of faculty members and students from the Georgia Institute of Technology are participating this week in the 2016 International Conference for High Performance Computing, Networking, Storage, and Analysis–better known as Supercomputing 2016 (SC16) in Salt Lake City.

“It’s hard to overestimate the role that high-performance computing and data science play in the modern world. With the proliferation of embedded sensors and computational devices across so many areas of human activity, we’re no longer talking even about a tsunami of data— it’s bigger than that,” said Georgia Tech School of Computational Science and Engineering (CSE) David Bader.

[Click here to see GT Computing's new HPC video]

“Supercomputing gives us the ability to make all this data work for us, inform us, and teach us. Georgia Tech is exceptionally well-positioned to work with a variety of partners to leverage HPC to the fullest extent of its capabilities, from building hardware to developing software and algorithms for these new supercomputers, to finding and analyzing the data that’s everywhere around us. It’s an exciting time,” said Bader.

Sponsored by the IEEE Computer Society, and the ACM Special Interest Group on HPC (SIGHPC), the SC conference attracts leading scientists, engineers, researchers, educators, programmers, system administrators, and developers from across the globe.

Along with staffing a booth (2543) during the exposition portion of the event, the Georgia Tech team will be participating in a number of ways at SC including paper and poster presentations, workshops, and panel discussions.

One of the papers being highlighted this year was actually presented at last year’s SC event. The paper written by students of CSE Professor Srinivas Aluru, “A Parallel Connectivity Algorithm for De Bruijn Graphs in Metagenomic Applications,” is being recognized with the first-ever ACM Results Replicated award and badge.

Another expected highlight this week at SC is the latest top 500 list of supercomputers. Known as the Graph 500, the bi-annual list is developed by a small cadre of recognized supercomputing experts, including Bader, and details the fastest and most powerful computers in the world.

One of the major challenges facing the further development of supercomputing is the looming end of Moore’s Law. To meet this challenge, SC is organizing the 1st International Workshop on Post-Moore Era Supercomputing (PMES). Bringing their expertise and insight to the inaugural workshop are Professor Tom Conte and Associate Professor Rich Vuduc.

Additional Georgia Tech participation includes:

Nov. 15 – 13th Graph500 List – David Bader

Nov. 15 – Students@SC Panel – Experiencing HPC for Undergraduates: Introduction to HPC Research – Edmond Chow

Nov. 16 – Panel – Post Moore’s Era Supercomputing in 20 Years – Tom Conte

Nov. 17 – Paper Presentation - A Parallel Algorithm for Finding All Pairs k-Mismatch Maximal Common Substrings – Sharma V. Thankachan, Srinivas Aluru

Nov. 17 – Paper Presentation – Performance Analysis, Design Considerations, and Applications of Extreme-Scale In Situ Infrastructures – Greg Eisenhauer, Matthew  Wolf, Suresh Menon, Reetesh Ranjan