SciNet GPU Workshop July 2010

From oldwiki.scinet.utoronto.ca
Jump to navigation Jump to search

On July 14th, SciNet held a well attended one-day workshop on GPU computing, bringing together SciNet users with interest in GPGPU (General Purpose computing on Graphics Processor Units), to exchange information and experiences. Representatives of NVidia and Software Carpentry took part as well.

Eight researchers from different departments at the University of Toronto gave short talks, followed by a presentation from Software Carpentry and a number of technical presentations from NVidia. Click on the titles to get the pdf or ppt files.


Research Talks

  • Abdul Mroue (Canadian Institute for Theoretical Astrophysics)
  • Teng Joon Lim (Electrical and Computer Engineering)
  • Jonathan Sievers (Canadian Institute for Theoretical Astrophysics)
    GPU correlation


Industry Presentations

  1. Introduction to Tesla GPU programming
  2. Advanced CUDA C
  3. Implementing 3D finite difference codes on the GPU
  4. CUDA development tools


SciNet GPU Resources

Note that SciNet currently has a Accelerator Research Cluster available to selected users for testing, code development, and evaluation purposes, but not as a production cluster. Contact us at support@scinet.utoronto.ca if you are interested.