On 9/19/07, Mark <[EMAIL PROTECTED]> wrote: > Timothy Normand Miller wrote: > > On 9/19/07, Farhan Mohamed Ali <[EMAIL PROTECTED]> wrote: > >> What kind of log function are you thinking about? Just an integer log? > > > > Integer log is too easy. :) > > > > For FP log, has this already been evaluated as a possibility? > > http://perso.ens-lyon.fr/florent.de.dinechin/recherche/publis/2005-Asilomar.pdf > http://www.ens-lyon.fr/LIP/Arenaire/Ware/FPLibrary/
I haven't looked at them, but the PDF looks like it might be an excellent starting point. With regard to other people's code, one of the issues to consider here is licensing. It's not necessary for all code discussed here to be given to Traversal. But Traversal was created to be entrusted with the ability to relicense code commercially. It's important that we generate revenue that we and the OHF can use to fund more development and manufacturing. So if we're looking at someone else's code, we won't accept it into the main OGP SVN tree unless Traversal has all appropriate permissions. (We could have a subtree where lots of Free code is hosted, but Traversal won't touch it.) > As an aside, is it one of the project requirements that all HDL be Verilog? It's a de facto requirement. It's important to be able to link lots of different components together, but most synthesizers won't let you mix languages. I'm not going to condone a debate on the relative merits of the languages, although I am aware of various advantages and disadvantages of each. My personal approach to chip design, however, is to make the language irrelevant by having in your mind the logic you want to design and using the language only as a superficial means of expression. That being said, people are free to translate designs between languages, just as they're free to translate our documents into other human languages. In fact, it's encouraged that proofs of concept be written in software languages (Ruby, C, C++, Perl, etc.) so that the math can be ironed out first. -- Timothy Normand Miller http://www.cse.ohio-state.edu/~millerti Open Graphics Project _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
