Has Task Creation performance improved significantly in .NET 4.5?

Question

Hi There,

I would like to know what I will gain by using .NET 4.5 for scenarios that I use TPL especially when I create new tasks.

Bryan

Answer 1

TPL allows your program use System resources effectively. If you have set of operations that can be performed parallely, then TPL automatically schedule those operations in a multiprocessor system. See that you could do the same thing without TPL but you would require to write large code. So, by using TPL your code not only becomes small, but also becomes efficient.

http://blogs.msdn.com/b/pfxteam/archive/2009/10/06/9903475.aspx

http://msdn.microsoft.com/en-us/library/dd537609(v=vs.110).aspx

---

Please mark this post as answer if it solved your problem. Happy Programming!

Answer 2

Hi Bryan,

TPL is designed to make it much easier to write managed code that can automatically use multiple processors. Using the library, you can conveniently express potential parallelism in existing sequential code, where the exposed parallel tasks will be run concurrently on all available processors. Usually this results in significant speedups.

Yo can find more information here:

http://msdn.microsoft.com/en-us/magazine/cc163340.aspx

---

Regards, http://shwetamannjain.blogspot.com

Answer 3

Hi There,

I would like to know what I will gain by using .NET 4.5 for scenarios that I use TPL especially when I create new tasks.

Geof

The TPL in general has had many improvements in .NET 4.5.  I recommend reading the TPL Performance Improvements article referenced here: http://blogs.msdn.com/b/pfxteam/archive/2011/11/10/10235962.aspx

It goes into detail about many of the improvements that come with 4.5, especially in regards to performance of the TPL.

 

 

---

Reed Copsey, Jr. - http://reedcopsey.com
If a post answers your question, please click "Mark As Answer" on that post and "Mark as Helpful".

Answer 4

Hi Sweta and Adavesh,

Thanks for your answers. My question is not about "what is TPL?", my question is: 

Has Task Creation performance improved significantly in .NET 4.5 comparing to the older .NET framework versions such .NET 4.0 ?

Regards,

Bryan

Answer 5

Bryan,

You can write the following code and measure the performance by yourself:

public static Tuple<long, long> CreateTasks(int ntasks)

{

    Task[] tasks = new Task[ntasks];

    Stopwatch sw = new Stopwatch();

    Action action = () => { };

    long startBytes = GC.GetTotalMemory(true);

    sw.Start();

    for (int i = 0; i < ntasks; i++) tasks[i] = new Task(action);

    sw.Stop();

    long endBytes = GC.GetTotalMemory(true);

    GC.KeepAlive(tasks);

    return Tuple.Create(sw.ElapsedMilliseconds,endBytes-startBytes);

}

You will 10~20 percent performance increase in average on different machines.

Regards,

Amir Ahani (MCSD.NET)

---

Answer 6

Thanks Amir and Reed.

I wrote the following code and I run it for both .NET 4.0 and .NET 4.5 on my machine, but surprisingly I receive the following results:

Test 1 -

In .NET 4.0, it takes 3351 Milli seconds, and in .NET 4.5, it takes 3570 Milli seconds

Test 2-

In .NET 4.0, it takes 3421 Milli seconds, and in .NET 4.5, it takes 3551 Milli seconds

namespace ConsoleApplication1
{
    class Program
    {
        static void Main(string[] args)
        {
            CreateTasks(10000000);
        }
        public static Tuple<long, long> CreateTasks(int ntasks)
        {
            Task[] tasks = new Task[ntasks];
            Stopwatch sw = new Stopwatch();
            Action action = () => { };
            long startBytes = GC.GetTotalMemory(true);
            sw.Start();
            for (int i = 0; i < ntasks; i++) tasks[i] = new Task(action);
            sw.Stop();
            long endBytes = GC.GetTotalMemory(true);
            GC.KeepAlive(tasks);
            Console.WriteLine(" {0} .... {1}  ", sw.ElapsedMilliseconds, endBytes - startBytes);
            return Tuple.Create(sw.ElapsedMilliseconds, endBytes - startBytes);

        }
    }

Please let me know what I am missing.

Thanks,

Bryan

Answer 7

Thanks Amir and Reed.

I wrote the following code and I run it for both .NET 4.0 and .NET 4.5 on my machine, but surprisingly I receive the following results:

Test 1 -

In .NET 4.0, it takes 3351 Milli seconds, and in .NET 4.5, it takes 3570 Milli seconds

Test 2-

In .NET 4.0, it takes 3421 Milli seconds, and in .NET 4.5, it takes 3551 Milli seconds

namespace ConsoleApplication1
{
    class Program
    {
        static void Main(string[] args)
        {
            CreateTasks(10000000);
        }
        public static Tuple<long, long> CreateTasks(int ntasks)
        {
            Task[] tasks = new Task[ntasks];
            Stopwatch sw = new Stopwatch();
            Action action = () => { };
            long startBytes = GC.GetTotalMemory(true);
            sw.Start();
            for (int i = 0; i < ntasks; i++) tasks[i] = new Task(action);
            sw.Stop();
            long endBytes = GC.GetTotalMemory(true);
            GC.KeepAlive(tasks);
            Console.WriteLine(" {0} .... {1}  ", sw.ElapsedMilliseconds, endBytes - startBytes);
            return Tuple.Create(sw.ElapsedMilliseconds, endBytes - startBytes);

        }
    }

Please let me know what I am missing.

Thanks,

Bryan

This isn't that good of a test, as you make a lot of Task instances, but don't actually start or use them....

A pure construction test like this is going to be fairly useless for real-world performance.  (Also, I suspect you weren't running outside of VS - those numbers sound like the test host was attached...)

---

Reed Copsey, Jr. - http://reedcopsey.com
If a post answers your question, please click "Mark As Answer" on that post and "Mark as Helpful".