Abstract Detail

Systematics Section / ASPT

Reeves, Aaron [1], Fisher, Ward I. [2], Simmons, Mark P. [1].

Atriplex: A simple-to-use platform for general-purpose grid computing.

Many contemporary phylogenetic and population genetic studies are computationally intensive: it is not unusual for an analysis to consume hundreds or thousands of hours of computing time. Relatively few investigators have access to dedicated computational resources to carry out such studies due to the expense and administrative difficulties associated with establishing and maintaining such facilities. An alternative approach, which may be used to make such studies more feasible and economical, would be to employ many ordinary personal computers (e.g., in offices or campus computer labs) when they are otherwise idle. A number of specialized systems have been developed for this purpose, but most require applications that have been specifically written for, or modified for use with, a particular system or application. We have developed and deployed Atriplex, a client/server platform for grid computing. Atriplex clients are standard multipurpose desktop computers that, when otherwise unused, carry out tasks and analyses assigned to them by an Atriplex server. Atriplex may be used to control a very wide array of existing or new computer programs, such as PAUP* and MrBayes. There is no need to rewrite or reengineer applications specifically for Atriplex, unlike other systems for distributed or grid computing. Atriplex is simple to use, and requires little specialized administration. Atriplex is currently available for Windows operating systems, although it is designed to be cross-platform compatible, and could be relatively easily adapted for Linux/Unix or Macintosh operating systems.

1 - Colorado State University, Department of Biology, Fort Collins, Colorado, 80523-1878, U.S.A.
2 - Colorado State University, Department of Computer Science, Fort Collins, Colorado, 80523, U.S.A.

Keywords:
grid computing
distributed computing
PAUP*
MrBayes
simulations
computationally intensive analyses.

Presentation Type: Poster
Session: 32-128
Location: Special Event Center (Cliff Lodge)
Date: Tuesday, August 3rd, 2004
Time: 12:30 PM
Abstract ID:373