COMING UP:

• Wed May 4, 12:00 noon: SNUG (SCINET USERS GROUP) MEETING
  • TechTalk by Peter Colberg (Chemical Physics Theory Group)

    “Scripting HALMD with Lua and Luabind”

    Lua is a lightweight, embeddable scripting language. Luabind is a library that exposes C++ classes and functions to Lua. HALMD, a highly accelerated large-scale molecular dynamics package for GPUs, uses Lua and Luabind to couple a set of high-performance C++ and CUDA modules in a flexible and extensible manner.

  • User discussion
  • Pizza!

  Sign up if you are planning to come, at https://support.scinet.utoronto.ca/courses/?q=node/25

  • MAY 9-13: 5 DAY PARALLEL PROGRAMMING COURSE

  The goal of the workshop is to enable young researchers already experienced in scientific computing to leave with the knowledge necessary to begin writing the parallel codes needed for their research. The workshop will be a mix of lectures and immediate feedback on practical assignments, designed to ensure that students leave with significant experience in both OpenMP and MPI, two of the standards for parallel computing today. We aim for this workshop to be useful for graduate students and postdocs for a wide range of disciplines.

  The course will be given in a yet-to-be determined location on the St. George Campus.

  There will be NO FEE, but you are responsible to get the recommended BOOKS, and your own COFFEE and LUNCH.

  There is LIMITED AVAILABILITY, so please only register if you are serious about attending. Registration and further details (topics, books, location) are (or will be) available at

  https://support.scinet.utoronto.ca/courses/?q=node/23

• May 30-June 3: ONTARIO SUMMER SCHOOL ON HIGH PERFORMANCE COMPUTING

  This Summer School is jointly sponsored by SHARCNET, SciNet and HPCVL — your Ontario HPC consortia. It is held at the Sheridan Institute of Technology and Advanced Learning in Trafalgar, Oakville, Ontario.

  This annual, intensive education event offers a week-long program of hands-on courses covering introductory HPC materials, programming for distributed systems (MPI) and shared memory environments (OpenMP), GPU programming, software best practices, debugging and much more.

  To maximize the ability for persons from the widest range of SHARCNET, SciNet and HPCVL institutions to attend, a subsidized rate will be available for those traveling a reasonable distance which includes up to 4 nights accommodation in the on-site residence, and a limited number of travel subsidies will be available, reserved for those traveling very long distances (Lakehead for example).

  Please watch the website (http://ss2011.sharcnet.ca) over the coming days as more information is put in place, and registration opens. Note that spaces are limited so register early!

  If you have additional questions in the meantime, please e-mail ss2011@sharcnet.ca .

• Wed Jun 8, 12:00: SNUG MEETING

  Let us know (support@scinet) if you’re interested in giving a TechTalk.

  Sign up if you are planning to come, at https://support.scinet.utoronto.ca/courses/?q=node/26

• June 13-17: HIGH PERFORMANCE COMPUTING SYMPOSIUM 2011

  HPCS2011 is a multidisciplinary conference that focuses on research involving High Performance Computing and its application.

  This year it is given in Montreal, and focuses on HPC in medical science.

  The actual symposium is from June 15 to 17. There are very worthwhile tutorial sessions on the two days before.

  More info can be found at at http://2011.hpcs.ca

ADDED TO THE WIKI IN APRIL:

(All new wiki content below is listed and linked on the main page: https://support.scinet.utoronto.ca/wiki/index.php/SciNet_User_Support_Library#What.27s_New_On_The_Wiki)

• Slides and source code of the Scientific Modern Fortran class.
• Slides of the April 2011 SNUG on User-space modules and packages.
• Instructions on installing your own modules.
• Instructions on how to install your own Python Modules.

SYSTEM CHANGES:

• GPC: Two versions of Goto Blas were installed, a single and multi-threaded one. They can be loaded as modules “gotoblas/1.13-singlethreaded” and “gotoblas/1.13-multithreaded”, respectively.
• TCS: A temporary workaround for the development nodes tcs01 and tcs02 was in place, but things are back to usual.
• Accelerator Research Cluster (ARC): A 8-node GPU test-cluster has been setup with a total of 64 Nehalem CPUs and 16 GPUs (NVIDIA, Cuda capability 2.0). As with the TCS, you do not have access to this by default, but let us know if you want to do GPU computing. More information at https://support.scinet.utoronto.ca/wiki/index.php/GPU_Devel_Nodes
• SciNet is on Twitter! Our twitter feed can be found here: https://twitter.com/SciNetHPC

WHAT ELSE HAPPENED AT SCINET IN APRIL?

• A SNUG meeting was held on Apr 13, with a TechTalk by SciNet on “User-space modules and packages on SciNet”.
• On Apr 19, a 1-day course “Introduction to to Scientific Programming with Modern FORTRAN” was given.
• There was scheduled downtime on April 27 to make repairs and improvements to the cooling system.
• Problems with the TCS were largely remedied.
• A thank-you goes out to all the users that cleaned up a good deal of their scratch disk space upon our request, when scratch was reaching 98% utilization in early April!