On Mar 5, 2015, at 9:27 AM, Vasco Alexandre da Silva Costa <vasco.co...@gmail.com> wrote:
> I see. But double precision floating point math is a lot slower than single > precision on certain architectures like the lower end NVIDIA cards. e.g. the > GeForce GTX 780 Ti has 5048 SP GFLOPS but only 210 DP GFLOPS. Even on the > higher end NVIDIA cards the DP FLOPS are like a third of the SP FLOPS. The > rendering system should be tunable for either speed or accuracy. That kind of > complicates things. As a first approach we can work only on getting the > double precision to work but the performance will suffer a great deal. Absolutely understood, but it’s not just speed vs accuracy. If a ray that should mathematically hit an object reports a miss, we will produce incorrect analysis results, which could lead to incorrect decisions that affect manufacturing designs or even impact lives. Imagine a calculator that will report 2+2=[5,4,3] VERY quickly, much faster than one that reports 2+2=4 every time. Would that be very useful? Perhaps for some situations, but not for most math problems. That’s kind of our situation here. We don’t do raytracing for pretty pictures. Fast and wrong is useless for our purposes. If single precision can be made to match results, that will obviously be more desirable and faster. Frankly don’t care if ti's single or double precision, but the behavior of a ray that should hit or miss must be demonstrably mathematically correct or at least consistent with current behavior. 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
