ParalleX Computational Model

When Petascale Is Just Too Slow

2009-05-04T15:07:00.000-07:00

Click Here for the Article

Beowulf is Dead?

2009-04-14T22:59:00.000-07:00

A very interesting article written by Dr. Eadline for Linux Magazine

http://www.linux-mag.com/id/7263/

HPX Documentation has Arrived

2009-03-11T14:04:00.000-07:00

The full documentation for the HPX is finally here: http://px.cct.lsu.edu/HPX.html

ISC 2009: Keynote Presentation HPC Achievement and Impact 2009

2009-02-20T12:39:00.000-08:00

http://www.supercomp.de/isc09/Program/Wednesday-Keynote-Presentation-HPC-Achievement-and-Impact-2009

Dr. Thomas Sterling: Keynote Speaker

Dr. Thomas Sterling Keynote from SuperComputing Comvention

2009-01-26T22:24:00.000-08:00

https://hpc.sun.com/feed/httpplanetssuncomhpcgroupblogsfeedatom/thomas-sterling-puts-past-year-perspective

Dr. Thomas Sterling puts the year of 2008 and HPC in perspective...

PXI (ParalleX Interface)

2009-01-13T12:06:00.000-08:00

We are currently working on the first release of the PXI Specification. PXI (ParalleX Interface) is the application programming interface (API) for the ParalleX programming model. This is equivalent to the case of MPI serving as the API for the communicating sequential processes model. Like MPI, it is a low level programming methodology with the strengths and weaknesses that this implies. It provides the user with the full set of semantic constructs of the ParalleX model for maximum control and flexibility, but does so at a cost of significant programming effort required with the resulting possibility of errors.

We expect to release version 0.1 of the PXI specification in February. Meanwhile, if you have any questions or comments, please visit our website http://px.cct.lsu.edu

HPC and Cloud Computing

2008-12-08T23:08:00.001-08:00

Check out this very interesting article published in HPC Wire by two members of our group (Thomas Sterling and Dylan Stark). You may also want to read the response from the editor of HPC Wire.

Brief Overview

2008-12-04T11:21:00.000-08:00

The goal of the ParalleX (PX) model of computation is to address the key challenges of efficiency, scalability, sustained performance, and powe consumption with respect to the limitations of conventional programming practices (e.g., MPI). ParalleX will improve efficiency by reducing the average synchronization and scheduling overhead, improve utilization through asynchrony of work flow, and employ adaptive scheduling and routing to mitigate contention (e.g., memory bank conflicts). Scalability will be dramatically increased, at least for certain classes of problems, through data directed computing using message-driven computation and lightweight synchronization mechanisms that will exploit the parallelism intrinsic to dynamic directed graphs through their meta-data. As a consequence sustained performance will be dramatically improved both in absolute terms through extended scalability for those applications currently constrained and in relative terms due to enhanced efficiency achieved. Finally, power reductions will be achieved by reducing extraneous calculations and data movements. Speculative execution and speculative prefetching are largely eliminated while dynamic adaptive methods and multithreading in combination serve many of the purposes these conventionally provide.

For more information, please visit our website