All posts by jchilders

ASCR FOA: Machine Learning and Understanding for High Performance Computing Scientific Discovery

ASCR has announced a new FOA. Details HERE.

Announcement Number: DE-FOA-0001575
Post Date: 04/18/2016
Close Date: 06/21/2016
Notes:

  • Letter of Intent Due Date: Not Applicable
  • Pre-Application Due Date: 05/20/2016 at 5 PM Eastern Time
    • (A Pre-Application is required)
  • Encourage/Discourage Date: 05/25/2016 at 5 PM Eastern Time
  • Application Due Date: 06/21/2016 at 11:59 PM Eastern Time

SUMMARY

The Office of Advanced Scientific Computing Research (ASCR) in the Office of Science (SC), U.S. Department of Energy (DOE), invites Continue reading ASCR FOA: Machine Learning and Understanding for High Performance Computing Scientific Discovery

DOE Graduate Student Research (SCGSR) Program

NOTE: Applications for the 2016 competition are due by December 15, 2015 at 5.00pm Eastern Time.

The SCGSR program provides supplemental support to graduate students working towards their PhDs. It allows the students to spend 3-12 consecutive months at a DOE National Laboratory to work on their thesis research topic.  U.S. citizenship is required and the work is to be conducted in a set of Priority Research Areas that have been identified for the SCGSR competition. For each competition priority areas are identified so applicants should check what they are. More information can be found at the DOE Graduate Student Research (SCGSR) Program website. Continue reading DOE Graduate Student Research (SCGSR) Program

Future Exascale High Performance Computing Input

DOE, NSF, and NIH recently released a joint RFI to identify science research topics nd applications needing High Performance Computing capabilities 100 times beyond today’s performance. The National Labs have already been asked for this information; university input is being actively sought. The deadline is October 16.

For more information, see the Request for Information (RFI) on Science Drivers Requiring Capable Exascale High Performance Computing.

ASCR/HEP Exascale Requirements Review

The first of the DOE Scientific Computing Exascale Requirements Reviews conducted by DOE ASCR was held in June 10-12, 2015 in Bethesda, MD. This was a joint ASCR/HEP program and was organized on the HEP side by the HEP-FCE. The agenda and the presentations can be found at the review website. A report that contains planning material for the next decade, including a number of white papers and case studies, will be available shortly.

The Largest LHC Event Generation Jobs Ever Run

Thomas LeCompte and Taylor Childers put supercomputers to work for Energy Frontier experiments by running LHC event generation at massively parallel scales on Mira, the supercomputer hosted by Argonne‘s Leadership Computing FacilityMPI (Message Passing Interface) was used to make Alpgen, a leading-order event generator, run in parallel threads. This work used a discretionary computing allocation on Mira to start the process of scaling to the minimum job size on Mira (512-computing nodes). An ASCR Leadership Computing Challenge award of 50M core-hours on Mira enabled the production of ATLAS event generation samples in preparation for Run-II.

1.5M threads of Alpgen running on Mira for 10 minutes to produce 85M W+5jet events.
1.5M threads of Alpgen running on Mira for 10 minutes to produce 85M W+5jet events