On Mar 6, 2015, at 4:32 PM, benson chepkwony <bchepk...@att.net> wrote:
> I am interested in "optimizing NURBS ray tracing" or "coherent ray tracing
> optimization"
>
> You Mentioned that:
> "We do implement NURBS geometry ray tracing, but not using a high-performance
> data-coherent method."
> Do you have any "data-coherent method" in mind?
This is not a simple discussion. Data coherency is not an algorithm you apply.
You’ll need to do a lot of homework if you don’t already have experience with
writing high-performing cache-coherent code:
http://en.wikipedia.org/wiki/Cache_coherence
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.16.8749&rep=rep1&type=pdf
http://www.cs.utah.edu/~shirley/irt/RT06_Course_LOD.pdf
https://fenix.tecnico.ulisboa.pt/downloadFile/3779574461367/RealTimeRayTracing.pdf
http://cseweb.ucsd.edu/~ravir/whitted.pdf
The second one from 2001 is particularly relevant, but they’re all good
resources you should understand in depth if you’re going to propose a GSoC
project in that area. Expecting to learn it during the GSoC coding window will
not work.
> Lastly, Should I conduct Profiling to identify bottleneck and NURBS "Hot
> Spots" in order to tell where I should put focus on?
You should always profile when doing any optimization work. You’ll want to
profile to write a decent project proposal involving optimization.
Cheers!
Sean
------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the
conversation now. http://goparallel.sourceforge.net/
_______________________________________________
BRL-CAD Developer mailing list
brlcad-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/brlcad-devel
