View Single Post
Old 05-03-2017, 06:55 AM   #1
pdblood
Junior Member
 
Location: Pittsburgh, PA

Join Date: Apr 2011
Posts: 4
Default Get fast and free access to large-memory computing

If you require large shared memory compute resources for de novo assemblies or other memory-intensive tasks, you may want to consider requesting compute time on Bridges (https://psc.edu/bridges), a uniquely flexible supercomputing resource hosted at Pittsburgh Supercomputing Center (https://www.psc.edu) and made available free of charge to US investigators and their international collaborators through the National Science Foundation's XSEDE project (www.xsede.org).

Bridges has forty-two 3 terabyte nodes (with 64-80 cores each) and four 12 terabyte nodes (with 288-352 cores each), along with hundreds of 128 gigabyte nodes for distributed workloads. Each node also has terabytes of local RAID storage for excellent performance on data-intensive tasks. With so many large memory nodes available, researchers experience fast turnaround for their large memory jobs. Jobs are allowed to run for up to two weeks at a time, and even longer by special request. Many common bioinformatics packages are already installed, and support staff at PSC will install additional packages upon request.

Researchers at US institutions who are new to XSEDE can request a Startup allocation for up to 1,000 terabyte-hours on Bridges large memory resources. Larger allocations can be obtained through a quarterly peer-review process. US researchers can add their international collaborators to their XSEDE allocation.

For more information on obtaining access to Bridges, please see https://psc.edu/about-bridges/apply or feel free to contact me directly. We would be happy to help you get your pipelines up and running on Bridges.

Philip Blood
blood@psc.edu
Senior Computational Scientist
Pittsburgh Supercomputing Center
Carnegie Mellon University
pdblood is offline   Reply With Quote