Well, it's a spec. It'll be interesting to see how the implementation goes.
64-bit instructions seem like they may be a bit of a bandwidth hog. Ages ago, I read a paper that did a study of the effects of different regfile sizes on performance and other things in RISC processors. It depended on workload a bit, but over-all the optimal regfile size was 32, although 16 was nearly as good, while 64 suffered diminishing returns. SPARC has 128 (or more) physical registers, but that's because it has a sliding window for fast function calls. That sliding window, actually a circular register file, is something that designers have complained about because of challenges with the routing. I've always felt (without necessarily being fully informed) that the sliding window thing for SPARC was one of those "computer science" concepts that sounds good on paper but isn't necessarily much of a win in practice. On Fri, Jul 13, 2012 at 8:38 AM, Vasileios Anagnostopoulos <[email protected]> wrote: > The relevant blog post http://random.axman6.com/blog/?p=327 > > The referred library > > http://ittc.ku.edu/csdl/fpg/Tools/KansasLava > > unfortunately in Haskell > > Regards > _______________________________________________ > Open-graphics mailing list > [email protected] > http://lists.duskglow.com/mailman/listinfo/open-graphics > List service provided by Duskglow Consulting, LLC (www.duskglow.com) -- Timothy Normand Miller, PhD 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)
