Class meeting for 10-405 Computing with GPUs

From Cohen Courses
Revision as of 15:17, 15 January 2018 by Wcohen (talk | contribs) (Created page with "This is one of the class meetings on the schedule for the course Machine Learning with Large Data...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

This is one of the class meetings on the schedule for the course Machine Learning with Large Datasets 10-405 in Spring 2018.

Slides

Quiz

  • No quiz for today!

Things to Remember

  • The types of problems that can be parallized with map-reduce clusters versus GPUs
  • SIMD versus MIMD parallization
  • The actual process involved in using GPUs for ML problems:
    • Minibatches are generated by the CPU
    • Minibatch data is copied from CPU to GPU
    • Vectorized version of classifier update step is executed on the GPU
    • So the ML weights (the model) and one minibatch are stored on the GPU