Parallel Programming Languages Spring 2010

• Time and place: Mon 12-3pm, Shannon-164.
• Professor: Jens Palsberg (palsberg@ucla.edu).
• Motivation for the course: learn new programming abstractions that will be useful for programming multi-core computers, clusters, GPUs, and supercomputers. I hope the course can help kick-start new research projects.
• Reading list.
• Homework.
• Homework 1 deadline: Sun May 9, 23:59. Send your implementation to Jens Palsberg (palsberg@ucla.edu).
• Lectures:
  • Mon Apr 12: overview, historical perspective, the landscape of parallel computing research, parallelism, synchronization, data distribution; example: map-reduce.
  • Mon Apr 19: X10, examples, core language, semantics, type system.
  • Mon Apr 26: connection between functional programming and parallel programming; inversion generates divide-and-conquer parallel programs.
  • Mon May 3: optimistic parallelism requires abstractions; software behavior oriented parallelization.
  • Mon May 10: static analysis of parallel programs; may-happen-in-parallel analysis of X10 programs.
  • Mon May 17: streaming; sketching; sensing.
  • Mon May 31: language presentations:
    • Haskell: Peter + Thomas;
    • Titanium: Jørgen;
    • Erlang: Ubbe + Steffen;
    • CUDA: Surayya;
    • Open CL: Andreas;
    • NESL + ZPL: Ian + Mathias.