reimplement a loop using Parallel.For

Question

how to re implement the loop below using Parallel.For?

for (int i = 0; i < data.Length; ++i)
  {
      int cluster = clustering[i];
      for (int j = 0; j < data[i].Length; ++j)
          means[cluster][j] += data[i][j]; // accumulate sum
  }

getting better performance and speed up is the goal.

We don't outright do your work for you here. You need to show us what you've tried already. — cost
– cost, Commented Dec 22, 2014 at 21:57
How big are your arrays? Do you really need parallelism? (Addition is one of the cheapest instructions) — L.B
– L.B, Commented Dec 22, 2014 at 21:59

Reed Copsey · Accepted Answer · 2014-12-22 22:00:40Z

3

You can mostly just replace the outer loop. However, you need to take care with the setting, as you're setting values from multiple threads:

Parallel.For(0, data.Length, i => 
{
  int cluster = clustering[i];
  for (int j = 0; j < data[i].Length; ++j)
      Interlocked.Add(ref means[cluster][j], data[i][j]); 
});

However, this may not run any faster, and may actually run significantly slower, as you could easily introduce false sharing since everything is reading from and writing to the same arrays.

answered Dec 22, 2014 at 22:00

Reed Copsey

567k80 gold badges1.2k silver badges1.4k bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Araz Over a year ago

thank you; Interlocked.Add Method overloads not support double type, as my arraye named "data" is double. I read related post in this case but this is not clear for me how to exactly rewrite your suggested loop.

Araz Over a year ago

This loop is one of my bottleneck in my algorithm as I'm interested to manage it using parallelism to getting speed up. what is any better solution considering False sharing and caching?

Reed Copsey Over a year ago

@Araz Without interlocked.add, you'd need to do some other form of synchronization (potentially using locking). It might be worth rethinking the algorithm's design to separate out the work so it can be parallelized more easily.

Collectives™ on Stack Overflow

reimplement a loop using Parallel.For

1 Answer 1

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related