→ Slides doi:10.7490/f1000research.1112729.1
AuthorsAlexander J. Ropelewski
, Pittsburgh Supercomputing Center
Philip D. Blood, Pittsburgh Supercomputing Center
Robert Light, Pittsburgh Supercomputing CenterAbstract
The Pittsburgh Supercomputing Center's (PSC) new computational system Bridges, funded by the National Science Foundation (NSF), is available to U.S. academic researchers through NSF's XSEDE program. Bridges is a unique system that consists of a variety of specialized nodes including: compute nodes, GPU nodes, database nodes, webserver nodes and data transfer nodes. A unique feature of Bridges is that the compute nodes are tiered in terms of memory, containing either 128GB, 3TB, or 12TB of hardware-supported shared memory, which makes the system ideal for Galaxy workflows involving Next Generation Sequencing data.
In this talk we will discuss the history of Galaxy at the PSC and describe various Galaxy usage scenarios for Bridges. These scenarios include (1) a shared galaxy instance for users with XSEDE allocations, (2) private "virtualized" instances of Galaxy, and (3) back-end computational support for remote Galaxy instances. We will also discuss the system that we developed to authenticate and charge usage against specific user-selected projects.