In this project we study in parallel programming issues: languages, scalability. programmer productivity and performance tuning. The main vehicle for this study is the IBM XLUPC compiler
The main goal of the project is to research scalability issues for parallel programming. We are looking at all aspects of programming, from language design, to runtime system research, program analysis and understanding, and compiler optimizations.
Accomplishments
In the context of the IBM PERCS project (part of the DARPA HPCS program) we developed a compiler for the UPC language based on the state-of-the-art IBM XL compiler framework. The XLUPC compiler will become part of the IBM product offering in 2H2009/1H2010.
Our team won the HPC Challenge Class 2 Awards in 2005 and 2006. More information on the IBM submissions is available here.
Team
| Research | Toronto Lab | Students |
|---|---|---|
| Calin Cascaval | Kit Barton | Yili Zheng, Purdue University |
| George Almasi | Philip Luk | Montse Farreras, UPC Barcelona |
| Anthony Bolmarcich | Ettore Tiotto | Rajesh Nishtala, UC Berkeley |
| Kevin Lou | ||
| Jason Chu | ||
| Wei Li | ||
| Raymond Mak | ||
| Lawrence Hu |
Last updated 7 Jul 2008