bees@infoswx.UUCP (06/13/86)
> I think someone should run the Dhrystone benchmarks on a 32332 > machine otherwise, I'd be force to to conclude that the part really > wasn't that fast. > -- > Clif Purkiser, Intel, Santa Clara, Ca. > HIGH PERFORMANCE MICROPROCESSORS > {pur-ee,hplabs,amd,scgvaxd,dual,idi,omsvax}!intelca!clif Let's see Clif, by that 'logic' - "I think someone should run 4.2BSD or System V, Rel 2 on an 80386 machine, otherwise, I'd be forced to conclude that the part really can't run demand-paged virtual UN*X." Pretty silly, huh! :-) Gary Rima, National Semiconductor, Dallas, TX EVEN HIGHER PERFORMANCE MICROPROCESSORS :-) {convex!infoswx,nsc!nsc16}!tx-nsc!rima
brooks@lll-crg.ARpA (Eugene D. Brooks III) (06/15/86)
It took months to get the (Motorola, National, Intel) wars off of net.arch. When net.micro.ns32k came on line I thought that it would be interesting to subscribe, as we have a multiprocessor based on this chip set. I now see where the (M,N,I) wars went to... If you guys would spend less time fighting and more time designing we would have faster parts! Its time for the old ug command! ;-)
jpm@quad1.UUCP (John McMamee) (06/17/86)
> It took months to get the (Motorola, National, Intel) wars off of net.arch. > When net.micro.ns32k came on line I thought that it would be interesting to > subscribe, as we have a multiprocessor based on this chip set. I now see > where the (M,N,I) wars went to... If you guys would spend less time fighting > and more time designing we would have faster parts! Some people actually find the processor wars interesting, so long as the majority of postings contain real information rather than personal attacks. I think the net has to face the fact that people like to argue about the various CPUs. I don't propose creating a new newsgroup to carry such discussions. These discussions are usally triggered by new chip announcements so the logical place to have the discussion is in the newsgroup devoted to the particular line of processors (i.e. the 32232 discussion belongs in net.micro.ns32k, a 68020 discussion belongs in net.micro.68k, etc.). The volume REALLY isn't that great (I'm only getting 4 or 5 messages a day in net.micro.ns32k). -- John P. McNamee Quadratron Systems Inc. UUCP: {sdcrdcf|ttdica|scgvaxd|mc0|bellcore|logico|ihnp4}!psivax!quad1!jpm ARPA: jpm@BNL.ARPA
clif@intelca.UUCP (Clif Purkiser) (06/18/86)
> It took months to get the (Motorola, National, Intel) wars off of net.arch. > When net.micro.ns32k came on line I thought that it would be interesting to > subscribe, as we have a multiprocessor based on this chip set. I now see > where the (M,N,I) wars went to... If you guys would spend less time fighting > and more time designing we would have faster parts! > > Its time for the old ug command! ;-) Mr Brooks is right this discussion has regressed into another microprocessor war. I guess I am as guilty as anyone for starting it again. (Look at the good side the group was kinda of slow until the discussion started.) Anyway I'm prepared to stop since net.micro.ns32k is not about benchmarking. One final comment, when someone makes a claim like "processor X runs faster than processor Y" I believe that it perfectly reasonable to expect justification for the claim. The justification doesn't not have be it runs Sieve twice as fast or X does 5000 dhrystones Y does 3000 dhyrstones. It could be my graphics package took less time to run on X than Y. However the dhrystone does seem to be the most popular attempt to measure performance on network. It is therefore a reasonable request to ask someone to run it for a new computer system. Lies Damn Lies Benchmarks -- Clif Purkiser, Intel, Santa Clara, Ca. {pur-ee,hplabs,amd,scgvaxd,dual,idi,omsvax}!intelca!clif {Stamp Out Stupid Signatures}