SciNet News February 2011

February 4, 2011 in for_users, newsletter

COMING UP:

(Unless stated otherwise, all events take place at the SciNet Headquarters, Rm 235 of 256 McCaul St., Toronto)

  • Wed Feb 9, 12:00 noon: SNUG (SciNet Users Group) meeting

    • Two TechTalks:
      1. Seth Dworkin will talk about: “Computational Combustion: Toward the use of Sustainable and Alternative Fuels.”
      2. SciNet will talk about “How to Monitor Your SciNet Jobs and Their Efficiency.”
    • User discussion
    • Pizza!

    Sign up if you are planning to come, at https://support.scinet.utoronto.ca/courses.

  • Feb 22, 23: Science Illustrated

    A two-day workshop on the technical and compositional techniques for creating good visualizations of scientific data. More information, and registration forms, will be available soon at http://www.scienceillustrated.ca.

    Note: this event is not held by or at SciNet, but we are involved.

  • Mar 9: SNUG meeting (TechTalks by Abdul Mroue and Joachim Harnois-Deraps)
  • Mar 15: Introduction to Scientific Programming with C++ (1 day)
  • Apr 13: SNUG meeting (TechTalks TBA)
  • Apr 19: Introduction to Scientific Programming with modern FORTRAN (1 day)
  • May 9-13 5-day Parallel Programming course:

    OpenMP, MPI and more will be covered. The aim is for you to leave the course able to start parallelizing your code. There will likely be a fee for this one.

    See https://support.scinet.utoronto.ca/courses for details and registration on the above SNUGs and courses.

ADDED TO THE WIKI IN JANUARY:

  • How to request ‘Network Switch Affinity’ for GPC Ethernet jobs at runtime.
  • Slides of the January TechTalks by Sarah Rauscher and Bijia Pang.
  • Past SNUG TechTalks page.
  • Updated and improved SciNet User Tutorial.
  • How to use Co-array Fortran on the GPC.

SYSTEM CHANGES:

  • For groups who were allocated compute time in this RAC allocation round, the new RAPs took effect on Jan 17th.

    Users in those groups will automatically run under the new RAPs (unless they change this themselves on the SciNet portal site).

    For users that are new to this allocation system, the allocation is implemented by giving your runs priority in the queue (fairshare).

    Groups with increased disk space allocations will get access to their full, newly allocated space later in the year, as we ramp up the installation of new storage (already in progress).

  • GPC: Users can now request ‘Network Switch Affinity’ for GPC Ethernet jobs.
  • File system servers were reconfigured to improve performance and stability. File access should be better, especially for writing.

WHAT ELSE HAPPENED AT SCINET IN JANUARY?

  • A SNUG meeting with TechTalks from Sarah Rauscher and Bijia Pang was held on Jan 12, with a record breaking 27 participants!
  • An “Intro to SciNet” session was held on Jan 18.
  • There was scheduled downtime on Jan 19-20 for maintenance of the chiller.
  • A one-day “Intro to Parallel Programming was held” on Jan 27.