I was avoiding participation in this because I don't know ARMish things, but it sounds like there is even more reason to avoid FP on them. This is odd since they seem to be made for DSP, aren't they? Does anyone know anything definitive about why FP math would be 10x faster on the ATMEGA?
MS Xiaohui Liu wrote: > You can refer to an earlier discussion > <http://mail.millennium.berkeley.edu/pipermail/tinyos-help/2011-February/049855.html>. > > In general, float ops should be avoided on motes. > > On Tue, Mar 29, 2011 at 5:15 AM, Marco Chirico <[email protected] > <mailto:[email protected]>> wrote: > > Hi list, > > I am working on an algorithm which performs intensive floating point > operations (e.g. float matrix multiplications, roots, etc.) on the > sensors. > I found that if I deploy it on a micaZ/iris mote everything is ok > and the job takes a time, say, t. > If I try to use another kind of mote (e.g. iMote, shimmer), it takes > a time which is, in best cases, 10t. > > Is there anyone who had the same issue with float ops? > > Thank you in advance, > Marco > > _______________________________________________ > Tinyos-help mailing list > [email protected] > <mailto:[email protected]> > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help > > > > > -- > -Xiaohui Liu > > > ------------------------------------------------------------------------ > > _______________________________________________ > Tinyos-help mailing list > [email protected] > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
