Join us August 8–11, 2011 for the next workshop on Gordon, SDSC's new NSF-funded (five year, $20 million award) data-intensive supercomputer. Gordon will be open for allocation requests this fall for a January 1, 2012 production start.
The Get Ready for Gordon Summer Institute is targeting users with applications that can potentially take advantage of Gordon’s unique capabilities for high performance data-intensive computing. We encourage applications from researchers engaged in data-intensive science and data mining across a wide range of disciplines, including those that have not typically considered use of supercomputing resources. The range of domains can be broad, say, from astronomy, geosciences and genomics, to economics and linguistics.
Gordon has been especially designed to address data-intensive problems that cannot make effective use of the current generation of massively parallel supercomputers. For example, developers of database applications can benefit from the large amount of flash memory in the system. Applications that serve a wide research community, e.g., through the use of a science gateway, are also encouraged to apply.
Topics covered during the institute include:
- Overview of the Gordon architecture
- Using flash to improve I/O performance of data intensive applications
- Using vSMP for large memory applications
- Application profiling of data intensive applications
- Hands-on sessions using Gordon I/O nodes and Dash, the prototype of Gordon
- Database and data mining applications
- Breakout sessions for those who provide support for data intensive architectures like Gordon
- How to write a successful Gordon allocation proposal