February 27, 2014 – March 01, 2014

​

Location

Louisiana State University, Baton Rouge, Louisiana, USA

Organizers

Juana Moreno, Louisiana State University
Mark Jarrell, Louisiana State University
Timothy Germann, Los Alamos National Laboratory
Emanuel Gull, University of Michigan
Karol Kowalski, Pacific Northwestern National Laboratory
Thomas Sterling, Indiana University

​

Website: Application Development for Exascale Computing Website

Overview

The Center for Computation & Technology (CCT) at Louisiana State University is organizing its 19th Annual Mardi Gras Conference, February 27 - March 1, 2014, in Baton Rouge, Louisiana. We are requesting support to host an ICAM Workshop on Application Development for Exascale Computing on February 26th. We are requesting funds to support junior scientists attendance to this workshop and the following conference in the amount of $10,000. A small amount of $1,000 is requested to support the cost of lunch and refreshments for the workshop participants. We expect to be able to attract 30 to 40 junior scientists from ICAM branches. The Mardi Gras conference, now a biannual event, concentrates each year on a dierent computational theme of current relevance. This year our theme is Application Development for Exascale Computing, designed to introduce possible solutions to large scalability problems. Cutting edge research on emergent phenomena in nuclear physics, materials science, biological matter, sustainable energy, coastal modeling, and weather prediction to cite a few, requires simulations that scale with the system size. Therefore, scientists working on realistic modeling of all these phenomena are finding a barrier on how to scale their codes and how to eectively make use of current and future cyber-infrastructure. This year’s Mardi Gras conference is designed to introduce possible solutions to this scalability problem. A series of talks and workshops are planned where the creators of modern runtime systems are put into direct contact with application developers providing an avenue for application developers to hear about innovative, new approaches to computation. Likewise, runtime system groups will educate developers on their capabilities.