patwa@ricks.enet.dec.com (My name is...) (11/07/90)
Hi, I am looking for technical literature on IBM RS/6000 architecture, namely journal or conference papers. Any survey paper that compares IBM RISC with MIPS, SPARC etc. would also be helpful. Has anyone recently seen such papers or read articles in any of the IEEE magazines (Micro, Computer...)? Your help is highly appreciated. Thanks, Nital P. patwa patwa@ricks.enet.dec.com ------------------------------------------------------------------------ Disclaimer: This is for my personal use only. ------------------------------------------------------------------------
andras@alzabo.uucp (Andras Kovacs) (11/09/90)
'IBM Research Journal' Vol 34 #1 (Jan 1990) has lots of info about the RS/6000 both hardware and software aspects. -- Andras Kovacs andras@alzabo.UUCP
nazief@m.cs.uiuc.edu (11/12/90)
People at Los Alamos ran some benchmarks on RS/6000 and compared their performances with those obtained from Sun4, DEC-3100 and a couple of vector machines. The results were published in their tech. report # LA-11831-MS, "Los Alamos Experiences with the IBM RISC SYSTEM/6000 Workstations". --Bobby Nazief
ndoduc@framentec.fr (Nhuan Doduc) (11/12/90)
In <2003@mountn.dec.com> patwa@ricks.enet.dec.com (My name is...) writes: >I am looking for technical literature on IBM RS/6000 architecture, >namely journal or conference papers. The best brochure I heartily recommend is: IBM RISC System/6000 Technology, Marmata Misra (editor), SA23-2619, 1990 (same quality as an old companion: IBM RT Personal Computer Technology, Frank Waters (editor), SA23-1057, 1986) and the best "magazine" is: IBM Journal of Research and Development, no. 1, vol. 34, 1/90 May I wish other manufacturers to "clone" thoses publications when they describe their processors and systems >Disclaimer: This is for my personal use only. >------------------------------------------------------------------------ this is also my personal appreciation only --nh Nhuan DODUC, Framentec-Cognitech, Paris, France, ndoduc@framentec.fr or ndoduc@cognitech.fr, Association Francaise des Utilisateurs d'Unix, France, doduc@afuu.fr
my@dtg.nsc.com (Michael Yip) (11/13/90)
nazief@m.cs.uiuc.edu mentioned that "people at Los Alamos ran some
benchmarks on RS/6000 and compared their performances with" a few
other machines. Can someone on the Net please post those performace
comparison summary to the Net. Thanks.
-- Mike
my@dtg.nsc.com
PS: Someone also mentioned that the RS/6000 architecture will be
very hard to shink or go up in clock speed. Is that true?
And why?jackk@shasta.Stanford.EDU (jackk) (11/13/90)
In article <1509@frapper.nsc.com> my@frapper.UUCP (Michael Yip) writes: > >PS: Someone also mentioned that the RS/6000 architecture will be > very hard to shink or go up in clock speed. Is that true? > And why? No. Proof by counterexample: This has already happened. The RS/6000 model 550, announced last week (See EE TImes and the Wall Street Journal), runs at 41+ MHz, and has a SPECmark of 54.3. Jack
rro@debussy.cs.colostate.edu (Rod Oldehoeft) (11/25/90)
In article <3300212@m.cs.uiuc.edu> nazief@m.cs.uiuc.edu writes: > >People at Los Alamos ran some benchmarks on RS/6000 and compared their >performances with those obtained from Sun4, DEC-3100 and a couple of vector >machines. >The results were published in their tech. report # LA-11831-MS, "Los Alamos >Experiences with the IBM RISC SYSTEM/6000 Workstations". A similar report is in the proceedins of Supercomputing '90, just past: "Performance Evcaluation of the IBM RISC System/6000: Comparison of an Optimized Scalar Processor with Two Vector Processors" Margaret L. Simmons and Harvey J. Wasserman Computer Research Group Computing and Communications Division Los Alamos National Laboratory Los Alamos, NM 87545 ABSTRACT RISC System/6000 computers are workstations with a reduced instruction set processor recently developed by IBM. This report details the performance of the 6000-series computers as measured using a set of portable, standard-Fortran, computationally-intensive benchmark codes that represent the scientific workload at the Los Alamos National Laboratory. On all but three of our benchmark codes, the 40-ns RISC System was able to perform as well as a single Convex C-240 processor, a vector processor that also has a 40-ns clock cycle, and on these same codes, it performed as well as the FPS-500, a vector processor with a 30-ns clock cycle.
rcomm@chudich.co.rmit.oz (Mark Makies) (03/19/91)
Excuse my ignorance, but I am interested in information about the IBM RS/6000. There aws reference made to it in a spec listing and it faired rather well against other processing elements. Any info will be appreciated, please email, thanks, mm... Mark Makies Digital Systems and Computer Engineering Group Royal Melbourne Institute of Technology, Australia email: rcomm@chudich.co.rmit.oz.AU phone: +613 660 2090