Charm++ parallel programming system

["Kale, Laxmikant V" <kale AT illinois.edu>]

Projections is the native performance tool that comes with Charm++ and is
great for visualizing such data.
I assume someone else will answer your question in more detail as well.

But: are you looking for a single-core (or single-node) data for thread
creation overhead? My guess is it is around or below 1us for context
switching (essentially saving register state). It an be optimized further
(by explicitly saving a partial live state), but in most applications we
deal with, that has not been an issue.

Also, remember that the more common method in Charm++ is to not use
threads that need a stack of there own, but simple entry methods (or
structured dagger). The overhead for them is even lower, and of the order
of a few function calls.
 

-- 

Laxmikant (Sanjay) Kale         http://charm.cs.uiuc.edu
<http://charm.cs.uiuc.edu/>
Professor, Computer Science     
kale AT illinois.edu
201 N. Goodwin Avenue           Ph:  (217) 244-0094
Urbana, IL  61801-2302          FAX: (217) 265-6582






On 2/28/13 12:07 PM, "Noel Keen" 
<ndkeen AT lbl.gov>
 wrote:

>Hello,
>
>I'm trying to get some perf data on a charm++ app.  Looking for the number
>of threads created and the expense of the creation/deletion, etc.  I'm
>trying to build charm++ with TAU on a Cray XT with limited success.
>But if someone knows anything about getting this type of perf data,
>let me know.  Thanks
>
>Noel
>_______________________________________________
>charm mailing list
>charm AT cs.uiuc.edu
>http://lists.cs.uiuc.edu/mailman/listinfo/charm