about System: currentTimeMillis()

[kgcan2000]

hello,
i am new in java,
i wrote a java program, my program is multiplication of two big matrixes and i used 10, 50 ,200, 1000 treads in it to execute. but when i computed the endtime-starttime of my program for each number of threads, i had increasing numbers according to the increasing number of threads.
i used "currentTimeMillis()".
for example:
10 threads==> 266
200 threads ==> 2484
1000 threads ==> 12549
why did i meet this result,
my computer in not multicore and i was expecting decreasing numbers according to increasing number of threads?

[Fubarable]

Why do you think that multiple threads decrease time? And how are you measuring time?

[kgcan2000]

first, thank you for you interested my problem,
i dont know if it is true, but i had an expectation that if we used multiple threats, the computation would finished earier because we divided the jop to parts, and
i used:

  " long startTime = System.currentTimeMillis();
           mainclass ext=new mainclass();
           ext.islemler(); 
    long endTime = System.currentTimeMillis();
    System.out.println(endTime-startTime); "

in detail,

islemler is:

public void islemler(){
    	"for(int x=1;x<=1000 ;x++)
/*    1,10,50,100,200,500,1000==> thread     number, we can change it if we want to do it by 1,10,50,100,200,500,1000 threads by replacing these numbers, here now there is "1000".*/
    	{      ThreadClass t;Thread a; 
    		     t=new ThreadClass(x,1000);
    			 a=new Thread(t);
    			 a.start();
    			
    	 }

[Fubarable]

Multi-threading per se cannot speed up computer processing, since each thread shares a bit of processor time and will slow the other down. The exception to this is with a multiple processor environment if the problem can be split into parallel non-dependent processes, and even then, there's no guarantee that Java and the OS will run things on different processors.

[kgcan2000]

thank you :)