Dynamic looping in CUDA

Hi all,

Is dynamic looping supported in CUDA/G80?

By dynamic looping I mean in the kernel code, you have some code snippet like this:

 ...

   

for( int i=0; i<dynamicLoopSize; i++)

{

 ... 

}

...

Here dynamicLoopSize varies for each thread.

Any suggestion/idea would be appreciated!

Cyrus

Yes, this is supported.