C# Programming Performance Tips - Part One

You can read all the C# performance tips from the following links,

C# has a total of 10 overload methods.

Most of the developers adopt the below approach.

string str = "Akshay|Patel";
Stopwatch s1 = new Stopwatch();
s1.Start();
string[] temp1 = str.Split('|');
Console.WriteLine(s1.ElapsedTicks.ToString());

Let’s change the approach; i.e., rather than passing the character directly, let us create an array of characters and pass them as array elements.

Stopwatch s2 = new Stopwatch();
s2.Start();
string[] temp = str.Split(new char[] {
'|'
});
Console.WriteLine(s2.ElapsedTicks.ToString());

Run the application and compare the execution time.

The result suggests adopting the second approach to save the execution time.

6 Comments

Join the conversation! Your thoughts help the community grow.

Dinas Ridas1y ago
Just swap s1 and s2 sections, and you will see, that 1st is 8 times faster than second
Hung Vo7y ago
I think they are not too much difference, try this code static void Main(string[] args) { var rawValue = "hello,world"; Stopwatch s1 = Stopwatch.StartNew(); for (int i = 0; i < 1000000; i++) { rawValue.Split(','); } s1.Stop(); Console.WriteLine(s1.ElapsedTicks); Stopwatch s2 = Stopwatch.StartNew(); for (int i = 0; i < 1000000; i++) { rawValue.Split(new char[] { ',' }); } s2.Stop(); Console.WriteLine(s2.ElapsedTicks); Console.WriteLine("-------"); Console.ReadKey(); }
Zulqadar Idrishi7y ago
No I don't think so, I tested it. The only different is when you put second method to first It will take much time and The second method will take less time. If you think second approach is taking less time then execute only the second one and see the magic.
- Akshay Patel7y ago
  I do agree that sometimes you get different numbers with single iteration, with thousands of iterations, with huge strings, time is just relative measurements. But when you check the actual implementation of split method, compilers will interpret a single character as a single-element character array. So when we pass char array itself then we can avoid that interpretation part: public string[] Split(char separator, int count, StringSplitOptions options = StringSplitOptions.None) { return SplitInternal(new ReadOnlySpan<char>(ref separator, 1), count, options); } and from memory point of view also we can check with dotmemory and you can observe the difference.
Rajesh Patel7y ago
Its clearly visible on attached image, when we use str.Split('|'), it call Split method with parameter [params], this results in an array allocation even if no parameter is passed in for the params parameter
Etienne Yamsi7y ago
But why?? I don't understand. Can you explain what's happening in background ? ?