Adding two arrays asynchronously

I am fairly new to asynchronous programming; I am developing a program that deals with numbers in different bases and I want to make the addition method asynchronous.

Firstly, it computes the linear convolution and does the carrying later, thus every iteration of the for-loop is independent (Example):

(Little Endianness btw) [2,7,1] + [8,7,5] = [10,14,6] => [0,5,7]

The question is, can I, if there are threads available, make the addition process faster by having the iterations done at the same time in different threads, and how?

>Solution :

In case of huge arrays (threading has its own overhead), you can try Parallel, e.g. Parallel.For:

  int[] left  = ...
  int[] right = ...
  int[] result = new int[left.Length];

  ...

  Parallel.For(0, left.Length, i => result[i] = left[i] + right[i]);

Let’s have a look on the effect:

  int N = 100_000_000;

  int[] left   = new int[N];
  int[] right  = new int[N];
  int[] result = new int[left.Length];

  // To prevent garbage collection while testing
  GC.Collect(2);

  Stopwatch sw = new Stopwatch();

  sw.Start();

  // Parallel version
  //Parallel.For(0, left.Length, i => result[i] = left[i] + right[i]);

  // Standard for loop version
  for (int i = left.Length - 1; i >= 0; --i)
    result[i] = left[i] + right[i];

  sw.Stop();

  Console.Write(sw.ElapsedMilliseconds);

Outcome (.Net 6 IA-64, Realease build, Core i9, 6 cores)

  200 - parallel version
  500 - for loop version