Cilk++ processor oblivious

Cilk++ processor oblivious

I have done some of the analysis of the Cilk++ examples.
I have a doubt .Generally as per the presentation it says that cilk++ is processor oblivious.Does it mean that it does not map the architecture the processor is using,whether being a heterogenous or homogenous multicore system.Thus it will almost give the same result in any comp

Secondly whether it works only at OS level to show the amount of parallelism that the particular program can create ,or does it even help the processors in distributing the tasks among themselves

Kindly elaborate

2 posts / 0 new
Last post
For more complete information about compiler optimizations, see our Optimization Notice.

By "processor oblivious" we mean that the same program should run well on any number of processors. The computation is not divided into a discrete set of processors (for example, 4 chunks of work for 4 processors).

Instead, the programmer is encouraged to designate enough parallelism so that the Cilk++ runtime system can balance the computation dynamically among however many processors are available.

Leave a Comment

Please sign in to add a comment. Not a member? Join today