Greetings! Robert Boyer <[EMAIL PROTECTED]> writes:
F> > What is the performance difference to acl2 from Shelter vs. regular mv? > > Obviously Matt will be able to give you a better, more > accurate, and more reliable answer, but I would guess that > the answer is that ACL2 running in GCL with Schelter mv, for > about ten years, has been about 10% to 20% faster than with > regular multiple-values. Enough of a performance > improvement that Schelter-mv was the default until about a > year ago, when Matt went after using regular Common Lisp > multiple values with a 'threads/parallelism' (non-GCL) > motivation for OpenMCL and SBCL. > 1) Are values intrinsically parallelizable, or rather, like let bindings, does the spec demand that they be processed left to right? I suspect the latter. 2) Is there any way to provide syntax for GCL's fork based parallelism which would match that of the others, at least approximately, to aid in comparisons using the same code? One thing I like about fork based is the (almost) transparent syntax -- no explicit locking, mutexing, etc. Take care, > Bob > > > > -- Camm Maguire [EMAIL PROTECTED] ========================================================================== "The earth is but one country, and mankind its citizens." -- Baha'u'llah _______________________________________________ Gcl-devel mailing list Gcl-devel@gnu.org http://lists.gnu.org/mailman/listinfo/gcl-devel
