June 16-20, 2013

Leipzig, Germany

Session Details

Name: Programming Models & Tools
Time: Thursday, June 20, 2013
9:00 AM - 10:30 AM
Room:   Hall 3
CCL - Congress Center Leipzig
Breaks:10:30 AM - 11:00 AM Coffee Break
Chair(s):   Bernd Mohr, JSC
Abstract:   The goal of the session "Programming Models & Tools" is to assess current practices and state-of-the-art in programming scientific and industrial HPC applications on parallel systems. With the advent of accelerators like GPUs or the Intel MIC and heterogeneous cluster architectures, but also with the start of the quest for exascale systems, many new parallel programming paradigms were introduced in the last years and all promise application developers to enable them to engineer efficient parallel programs more easily. The first two talks review the traditional style of HPC programming based on Fortran and MPI and report on current and future developments and standards in this area. The third presentation looks on emerging and currently heavily discussed approaches for programming accelerators. The fourth talk closes the circle and provides visions for programming future exascale systems.  
Presentations: The Message Passing Interface: MPI 3.0 & the Road to MPI 4.0
9:00 AM - 9:20 AM
    Martin Schulz, LLNL
Modern Fortran & the Coarray Programming Model
9:20 AM - 9:40 AM
    Roberto Ansaloni, Cray
CUDA Past, Present & Future
9:40 AM - 10:00 AM
    Ian Buck, NVIDIA
Programming Models for Next Generation HPC Systems
10:00 AM - 10:20 AM
    Kathryn O'Brien, IBM
Questions & Answers
10:20 AM - 10:30 AM
Program may be subject to changes.