PAR Lecture 12, Mon Feb 27

1   Parallel.ecse programs

  1. /parallel-class/cuda/checksum.cc shows a significant digits problem when you add many small numbers.

  2. sum_reduction.cu is Stanford's program.

  3. sum_reduction2.cu is my modification to use managed memory.

    Note how both sum_reduction and sum_reduction2 give different answers for the serial and the parallel computation. That is bad.

  4. sum_reduction3.cu is a mod to try to find the problem. One problem is insufficient precision in the sum. Using double works. However there might be other problems.

2   Stanford lectures

Continue with slide 29 of Lecture 6 parallel patterns 1, which presents some paradigms of parallel programming. These are generally useful building blocks for parallel algorithms.

Then do lecture_7.

Comments

Comments powered by Disqus