Grid Computing Platform
Expanding Research Capabilities
With Wharton's Grid and Cluster Computing Platform, Wharton faculty, doctoral students, and research colleagues can analyze vast quantities of data – dramatically expanding the scope and complexity of research they can conduct.
Complex asset allocation studies that used to take two weeks to process can be completed in a day. And while researchers formerly had to rely on representative samples of data, subject to errors of sampling constraints, with the Wharton platform they can analyze entire data sets.
Wharton researchers can now ask questions that would not have been asked before, and can revisit past research, reanalyzing all the data to see whether the results are consistent with earlier studies based on samples – or whether different conclusions should have been drawn.
Wharton's computing platform gives researchers the tools to:
- Analyze problems that require complex processing, significant disk space, and sufficient RAM to run large amounts of live data
- Run interactive or batch jobs
- Check the status of their jobs over the Web through the grid monitor
Technical Platform
Software:
- Redhat Linux
- Sun Grid Engine Enterprise Edition
- Intel C/C++/Fortran Compilers with IMSL libraries
- MPI and GNU C/C++/Fortran Compilers
- Mathematica and gridMathematica
- Matlab
- mySQL
- R
Hardware:
- Grid of 32 dual-CPU Sun Fire V210 servers
- 32 terrabyte NetApp disk appliance
For More Information
- Wharton Grid Project: http://grid.wharton.upenn.edu/
- Grid Forums: https://grid.wharton.upenn.edu/forum/tiki-forums.php
