Mark - I ran reproduce your ram troubles with the PDL code that's contained in your YouTube description. I suspect it has something to do with how one creates animations using TriD, and that repeatedly calling points3d. Specifically, I suspect (though haven't dug through the TriD code enough to confirm) that each time you call points3d, the data gets copied somewhere, and that copy never gets destroyed.
Chris, does my explanation sound plausible? David On Thu, Jan 3, 2013 at 2:06 PM, MARK BAKER <[email protected]> wrote: > > here is the pastable text ... Tested it on Ubuntu.. > both work.. posted below > > -Mark > > ------------------------------ > *From:* MARK BAKER <[email protected]> > *To:* "[email protected]" <[email protected]> > *Sent:* Thursday, January 3, 2013 8:56 AM > *Subject:* Re: [Perldl] review of sf.net bugs/features for PDL-2.4.12 > > I Have been noticing a problem with the ram usage with PDL > I wrote some code in PDL then in OpenCL and the OpenCL > code Doesn't use any ram while the PDL code slurps up > about .01 Gigabytes per second , which really starts to add > up ... > > here is the code for both ... > > OpenCL code > > > http://www.youtube.com/watch?v=Tm2-YCGvBvo&feature=share&list=UUPHFUHiwbpMqC8ONxEICCiQ > PDL code > > > http://www.youtube.com/watch?v=3LXUz_Qez4A&feature=share&list=UUPHFUHiwbpMqC8ONxEICCiQ > > > > --Mark > > > > Given the recent work on plotting libraries, I wonder if you, Craig, Dima, > and I might consider working on stronger support for PDL::Graphics2D? As > your work on the tracker indicates, there are quite a few other subsystems > that could use help, so perhaps we should focus some effort now on those, > and come back to the graphics end later. Furthermore, that'll be a pretty > big project, and I would prefer to hold off until this summer due to tuits > and timing. > > That having been said, if it would work better for you guys to hit that > now, I would be game. :-) > > David > > > On Wed, Jan 2, 2013 at 12:25 PM, Chris Marshall <[email protected]>wrote: > > I've just completed a review of the PDL Bugs and Feature > Requests tracker items outstanding at > http://sourceforge.net/tracker/?group_id=612&atid=100612 > and > http://sourceforge.net/tracker/?group_id=612&atid=350612 > with an eye to goals for an upcoming PDL release and what > would make a good target. > > I urge PDL users and developers to look over the existing > Bugs and Feature Requests. If any strike your fancy, please > feel free to step up and code. Also, if there is something > you think would be a good/useful addition to PDL, add an > item to the Feature Requests or bring it up on this list. > > Thanks much, > Chris > > _______________________________________________ > Perldl mailing list > [email protected] > http://mailman.jach.hawaii.edu/mailman/listinfo/perldl > > > > > -- > "Debugging is twice as hard as writing the code in the first place. > Therefore, if you write the code as cleverly as possible, you are, > by definition, not smart enough to debug it." -- Brian Kernighan > > _______________________________________________ > Perldl mailing list > [email protected] > http://mailman.jach.hawaii.edu/mailman/listinfo/perldl > > > > _______________________________________________ > Perldl mailing list > [email protected] > http://mailman.jach.hawaii.edu/mailman/listinfo/perldl > > > > _______________________________________________ > Perldl mailing list > [email protected] > http://mailman.jach.hawaii.edu/mailman/listinfo/perldl > > -- "Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it." -- Brian Kernighan
_______________________________________________ Perldl mailing list [email protected] http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
