Class meeting for 10-605 Computing with GPUs

From Cohen Courses
Jump to navigationJump to search

This is one of the class meetings on the schedule for the course Machine Learning with Large Datasets 10-605 in Fall_2017.



  • 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