Path: utzoo!attcan!uunet!husc6!cmcl2!nrl-cmf!ames!pioneer.arc.nasa.gov!eugene From: eugene@pioneer.arc.nasa.gov.arpa (Eugene N. Miya) Newsgroups: comp.arch Subject: Re: getting rid of branches Message-ID: <11306@ames.arc.nasa.gov> Date: 5 Jul 88 18:17:01 GMT References: <1941@pt.cs.cmu.edu> <3208@ubc-cs.UUCP> <1986@pt.cs.cmu.edu> <853@garth.UUCP> <822@l.cc.purdue.edu> Sender: usenet@ames.arc.nasa.gov Reply-To: eugene@pioneer.arc.nasa.gov.UUCP (Eugene N. Miya) Organization: NASA Ames Research Center, Moffett Field, Calif. Lines: 40 In article <822@l.cc.purdue.edu> cik@l.cc.purdue.edu (Herman Rubin) writes: >In article <853@garth.UUCP>, smryan@garth.UUCP (Steven Ryan) writes: >> CFT is supposed to provide a vectorisable conditional expression. FTN200 code >> to vectorise ifs may/may not be completed and released. > >Even better is to first construct vectors A1, which contains only the elements >of A for which bit[i] set, and A0, which has the elements for which bit[i] is >clear, compute fn1 on A1, fn2 only on A0, and use bit to merge the two result >vectors. I think it was Steve who asked what this type of conditional code looks like. And I also see that Herman responded. I have collected some of this type of information for a little net article on supercomputers and what should be taught in super-classes which one might not get but should.... (they are just some ramblings sort of like these). Anyway, I spoke to one of my officemate for his sage opinions and he gave me a couple of references by Paul Frederickson. Well! Paul just took a job at RIACS, so I just walked down the hall, and he has this paper in Parallel Computing in 1984 and a more recent version in 1987. George only told me PC84 and the author, so you can look up the topic. Anyway, these codes are basically Monte Carlo bomb (nuke) codes from places like LLNL, LANL, Sandia which model photon, neutron, and energy transport, those kinds of things. Herman's description is closer (dope vectors). Anyways, if you need a reference, Parallel Computing 1984, a quarterly, article by Frederickson, so you shouldn't have too much trouble finding this, or consult your nearest National Laboratory. "Mr. Fusion" 8-) "Plutonium are your corner drug store..." Another gross generalization from --eugene miya, NASA Ames Research Center, eugene@aurora.arc.nasa.gov resident cynic at the Rock of Ages Home for Retired Hackers: "Mailers?! HA!", "If my mail does not reach you, please accept my apology." {uunet,hplabs,ncar,decwrl,allegra,tektronix}!ames!aurora!eugene "Send mail, avoid follow-ups. If enough, I'll summarize." There seems to be a little bit of interest in the subject of conditionals.