SciNet News December 2013

December 9, 2013 in for_researchers, for_users, newsletter, Uncategorized

To keep users posted on what’s going on at SciNet, every month, we compile the new features, planned courses and events and accomplishments at SciNet.

Speaking of accomplishments, we are compiling user success stories and testimonials. If you want to share your SciNet-powered success stories, please let us know by emailing to support AT scinet DOT utoronto DOT ca.

Below you will find the salient points of what has happened or changed during the last month and what will happen in the near future. Among these is the start of the term-long graduate course in Scientific Computing, which can be taken for credit by physics, astrophysics and chemistry graduate students at the University of Toronto.

As always, details are available on the wiki or the SciNet education website.

Happy new year!

– The SciNet team

Events Coming Up

Unless stated otherwise, all events take place at the SciNet Headquarters, Rm 235 of 256 McCaul Street, Toronto. All events below are free for users but we ask that you sign up (“enroll”) on the education website.

  • Wednesday December 11, 2013, 10:30 am – 11:30 am

    INTRO TO SCINET

    A class of approximately 90 minutes where you will learn how to use the systems. Experienced users may still pick up some valuable pointers during these sessions.

    Participation counts towards the SciNet HPC Certificate.

    For more information and enrollment, go to the course website.

  • Wednesday December 11, 2013, 12:00 noon – 1:00 pm

    SCINET USER GROUP (SNUG) MEETING

    The SciNet Users Group (SNUG) meetings are every month on the second Wednesday, and involve (free) pizza, user discussion, feedback, and a one or two short talks on topics or technologies of interest to the SciNet community.

    This time, we will have

    • TechTalk:

      New and Existing Resources Available at SciNet

    • User discussion
    • Pizza!

    For more information and enrollment, go to the course website.

  • Winter 2014

    SCIENTIFIC COMPUTING COURSE

    Many computational projects start off with knowledge of the science you want to do, and with a bit of programming experience. It can be an arduous journey to get to a (maintainable) piece of code which you trust to compute the right thing. This course is aimed at reducing your struggle, and make you a more efficient computational scientist. Topics include well-established best practices for developing software as it applies to scientific computations, common numerical techniques and packages (so you don’t reinvent the wheel), and aspects of high performance computing.

    The course consists of three parts:

    • Part 1: Scientific Software Development & Design
    • Part 2: Numerical Tools for Physical Scientists
    • Part 3: High Performance Scientific Computing

    Each part consists of eight lectures of one hour.

    Note that these parts can be taken as “mini-courses” by astrophysics graduate students and as “modular courses” by physics and chemistry graduate students at the University of Toronto. (Feel left out? Talk to your graduate office!)

    Participation in parts 1 and 2 counts towards the SciNet Scientific Computing Certificate.

    Participation in part 3 counts towards the SciNet HPC Certificate.

    For more information and enrollment, go to the course websites for
    part 1, part 2, and part 3.

  • January 7, 9, 14, 16, 21, 23, 28, 30: 11:00 am – 12:00 noon

    SCIENTIFIC SOFTWARE DEVELOPMENT

    Lectures of Part 1 of the Scientific Computing Course

    Topics: C++, version control, make, modular programming, testing, debugging, profiling

    For more information and enrollment, go to the
    course website.

  • Wednesday January 8, 2014, 12:00 noon – 1:00 pm

    SCINET USER GROUP (SNUG) MEETING

    TechTalk: TBA

    For more information and enrollment, go to the course website.

  • Wednesday February 12, 2014, 12:00 noon – 1:00 pm

    SCINET USER GROUP (SNUG) MEETING

    TechTalk: TBA

    For more information and enrollment, go to the course website.

  • February 4, 6, 11, 13, 25, 27, and March 4 and 6: 11am-12noon

    NUMERICAL TOOLS FOR PHYSICAL SCIENTISTS

    Lectures of Part 2 of the Scientific Computing Course

    Topics: Modelling, floating point computations, random numbers and Monte Carlo, ODEs and Molecular Dynamics, linear algebra, fast Fourier transforms

    For more information, go to the course website.

  • Wednesday March 12, 2014, 12:00 noon – 1:00 pm

    SCINET USER GROUP (SNUG) MEETING

    TechTalk: TBA

    For more information and enrollment, go to the course website.

  • March 11, 13, 18, 20, 25, 27, and April 1 and 3: 11am-12noon

    HIGH PERFORMANCE SCIENTIFIC COMPUTING

    Lectures of Part 3 of the Scientific Computing Course

    Topics: parallel programming concepts, shared memory parallel programming with openmp, distributed memory parallel programming with mpi, hybrid programming.

    For more information and enrollment, go to the course website.

  • Wednesday April 9, 2014, 12:00 noon – 1:00 pm

    SCINET USER GROUP (SNUG) MEETING

    TechTalk: TBA

    For more information and enrollment, go to the course website.

  • Wednesday May 14, 2014, 12:00 noon – 1:00 pm

    SCINET USER GROUP (SNUG) MEETING

    TechTalk: TBA

    For more information and enrollment, go to the course website.

System News

  • SciNet network connection will be down on Monday 9 Dec from 7:00 am to 11:00 am. Jobs will continue to run but users will not be able to connect to any SciNet systems.
  • The user-contributed x86 system “Sandy” is now available for job submission by other SciNet users as well. Jobs would be scheduled with default priority, depending on node availability. This Sandybridge x86 cluster has 76 nodes with 16 cores and 64GB of RAM per node. For more information on the system and how to use it, see the wiki.
  • Access to the user-contributed GPU system “Gravity” can now be requested by users (similar to ARC). Gravity is a cluster of 49 nodes, each node with 12 cores, 32 GB, and two NVIDIA Tesla M2090 GPUs. The GPUs have CUDA capability 2.0, 512 CUDA cores and 6 GB of RAM. For more information on the system and how to use it, see the wiki.
  • BGQ: Python 2.7.3 installed, with numpy support.
  • GPC: ffmpeg, a movie encoder, installed as a module.
  • P7: cmake 2.8.8, as build system, installed.
  • GPC, P7: Visit 2.6.3, a visualization program, installed as a module.

Added to the Wiki

All new wiki content below is listed and linked on the wiki main page.

  • Slides and Recordings of (almost) all lectures of “Research Computing with Python”
  • Links to recent TechTalks
  • BGQ Software list updated

What Else Happened at SciNet in the Last Month?

  • November 5 – 28, 2013: Final four lectures of “Research Computing with Python”
  • November 13, 2013: SNUG Meeting with TechTalk on Molecular Motors by Peter Colberg
  • November 18 and 19, 2013: Remotely hosted HPCVL’s “Introduction to HPC”.
  • November 22 and 23, 2013: Two unrelated events, a power glitch and a fabric manager dying caused running jobs to get lost. Both occurred at night, but systems were back up around 10am in the morning.