ARCHIVED: On Gordon (SDSC), what are the "bigflash" compute nodes?

This content has been archived, and is no longer maintained by Indiana University. Information here may no longer be accurate, and links may no longer be available or reliable.

On the Extreme Science and Engineering Discovery Environment (XSEDE), Gordon (SDSC) has four compute nodes available with the "bigflash" property. Each of these nodes has access to a separate 4.4 TB solid-state drive (SSD) file system. The "bigflash" compute nodes are intended for jobs needing larger scratch space than the 280 GB available to Gordon's native (non-vSMP) nodes.

To run a "bigflash" job on Gordon, change your PBS node request line:

  #PBS -l nodes=number_of_nodes:ppn=16:native:bigflash

If your job needs more than one "bigflash" node, add this line:

 #PBS -v Catalina_maxhops=2

The latency to the SSDs is several orders of magnitude lower than that for spinning disk, which makes them ideal for user-level check pointing, staging of data files that are accessed multiple times through the course of a job, and applications that need fast random I/O to large scratch files.

For more about storage for jobs on Gordon, see Storage Overview in the San Diego Supercomputer Center (SDSC) Gordon User Guide.

For more about XSEDE compute, advanced visualization, storage, and special purpose systems, see the Resources Overview, Systems Monitor, and User Guides. For scheduled maintenance windows, outages, and other announcements related to XSEDE digital services, see User News.

This document was developed with support from National Science Foundation (NSF) grants 1053575 and 1548562. Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF.

This is document bcua in the Knowledge Base.
Last modified on 2018-01-18 17:23:19.