GPU621/DPS921

GPU621/DPS921 | Participants | Groups and Projects | Resources | Glossary

Please help make this page resourceful for all GPU621/DPS921 students to use!

Course Material

GPU621/DPS921 - Parallel Algorithms and Programming Techniques

• Large data and compute-intensive problems benefit from parallel solutions.  Modern hardware has parallel processing capabilities in varying degrees.  Students study a set of industry-standard parallel patterns and learn how to implement parallel algorithms on multi-processor accelerators, shared memory systems and distributed systems using a range of programming models.
• GPU621 Course Outline
• DPS921 Course Outline

Instructor

• Chris Szalwinski

External Links

• Course Web Site – Lecture Notes
• Course Web Site – Timeline

Workshops

• The workshops provide timely opportunities to implement some of the material covered during the lectures. Each workshop is graded and all submissions are through Moodle.
• Detail Specifications
  1. Platforms
  2. False Sharing
  3. Prefix Scan
  4. Convolution
  5. Threading Building Blocks
  6. Vectorization
  7. Data Decomposition
  8. Scatter Gather
  9. Fork Join
  10. Optimization
• Grading - The due date for each workshop is noted in MySeneca. The penalty for late submission is 50% of the workshop mark.

Evaluation

• Assignments and Presentation 30%
• Workshops 20%
• Test 20%
• Exam 30%

Resources

• Current Software Support
  • Matrix
    • Cilk Plus - /usr/local/gcc/gcc-cilk/bin/g++ - version 4.9.0 experimental
    • TBB - /usr/local/gcc/tbb - version 4.1 update 4 (20130613)
  • T2108
    • Matrix Image (as above)
    • CUDA 5.5 July 2013 - supports Visual Studio 2012

• Wikis
  • How To edit Wiki pages
  • How To edit Wiki Cheatsheet