Bioinformatics Petascale Pioneers Program

The Pawsey Supercomputing Centre seeks expressions of interest for pioneering projects in the field of bioinformatics to use the Pawsey resource Galaxy during the first six months of 2015. This follows on from the highly successful original Petascale Pioneers Program, and has been prompted in part by the overwhelming positive response to the Pawsey Petascale Bioinformatics Symposium.

Galaxy is a Cray XC30 supercomputer containing 134 compute blades, each of which has four nodes. The majority of the nodes (484) support two ten-core Intel Xeon E5-2690v2 ‘Ivy Bridge’ processors with 64 GB of memory, for a total of 9,440 cores delivering roughly 220 TFLOPS (peak) computing power. The remaining 64 nodes consist of a single Intel Xeon E5-2670 ‘Sandy Bridge’ host processor with 32 GB of memory plus an NVIDIA K20X ‘Kepler’ GPU accelerator with 6 GB memory, delivering an additional 90 TFLOPS (peak) computing power. The machine interconnect is Cray’s high-speed, low-latency Aries interconnect in the Dragonfly topology, and the machine is attached to a Cray Sonexion 1600 appliance with 1.3 PB of scratch storage space.

We anticipate awarding on the order of six projects a share of 12 million CPU-hours and Pawsey Supercomputing Team expertise to assist in transitioning to the supercomputing environment. In addition to helping to port codes to the unique Cray environment, we will provide each project with help in scaling team-developed codes or transitioning to new applications more suitable to the high-performance computing environment. In addition, near the beginning of the six-month period we will host a week-long mentorship session in Perth for all projects selected, with Pawsey staff and national and international experts assisting the project teams. For interstate teams, limited matching travel grants will be available.

To register your interest, please submit a short description of your project to [email protected] (questions may also be sent to this address). The submission must include the following:

• Name, address, affiliation, contact information
• Project title
• Project team
• Project science description
  • A description of the science problem your project is trying to address
  • A description of how your project would benefit from the unique resources available in a supercomputing environment (as opposed to a workstation or desktop computer)
• Project code descriptions, programming environment, and support
  • What code(s) do you plan to use? Describe your planned workflow.
  • What libraries/software does your project require?
  • How do you plan to accommodate for the walltime limits on Galaxy (expected to be 24 hours)? (e.g., checkpointing, running on larger processor counts to reduce walltime, etc.)
  • What storage does your project require? (scratch, short-term, long-term)
  • What support would you need from Pawsey in order for the project to be successful?
• Project outcomes
  • Desired science outcomes
  • Other anticipated outcomes (e.g., verification that workflow incorporating new code will work for your science; submitting a partner scheme allocation request)

This opportunity is available to researchers from iVEC partner institutions and those employed in other Australian universities and research institutions. PhD students may not apply as the principal investigator, however they might wish to discuss this opportunity with their supervisors who could in turn submit a proposal as principal investigator.

The deadline for expressions of interest is 5pm Western Australian Standard Time on Friday 12^th December 2014 and we expect to announce the successful projects shortly thereafter.