This is NeXT-vs-Sun3 in view mode; [Up]
Date: Sun 19-Oct-1989 17:13:51 From: Unknown Subject: NeXT vs Sun3 I am getting ready to purchase a workstation. I had pretty much settled on either a Sun 3/80 or the Sparc1 machines. I can get the 3/80 with 4mb of memory, 19" monitor, diskless for about $5000.00. The Sparc1 is about $7000.00 with a 17" monitor and 8mb of memory. For $10000.00, I can buy a NeXT machine with 8mb of memory, 17" monitor, 256mb optical disk, etc... The big questions I have about the system are these: 1) Is the optical disk too slow (for virtual memory paging,etc) ? (And do large applications take forever to load?) 2) Is the propietary display system going to hinder software 3rd party software that either I write, or that I want to get? This is especially critical now that I can run X/Open on my Sun 3/80 at work under the X/NeWS display postscript system. Maybe X/NeWS will be ported to the NeXT. Does anyone have any comments that would help me make my decision ? Thanks, -Jaz (jaz@icd.ab.com) >From: kean@nyssa.CS.ORST.EDU (Kean Stump)
Date: Sun 20-Oct-1989 18:10:40 From: Unknown Subject: Re: NeXT vs Sun3 I ran my cube at home for 3 months off an optical drive before I bought my nice big Wren V. Optical only is a wee bit slow; NeXT has remedied that as of last Tuesday, by distributing a 40 Mb scsi disk with each optical only cube for the kernel to swap to (called the "accelerator"). Same price as the original optical only cube (6500.00 base) and, NeXT is (according to the campus rep) shipping 40 Mb scsi drives to those who purchased an optical only system. kean Kean Stump, College of Oceanography Domain: kean@{cs,oce}.orst.edu Oregon State University, Corvallis OR 97331-5503 UUCP: kean@shatter.UUCP ****OSU knows nothing about my opinions. Absolutely nothing. So, don't ask.**** Diplomacy : The art of saying "nice doggy", until you can find a big rock. >From: nevai@gem.mps.ohio-state.edu (Paul Nevai)
Date: Sun 21-Oct-1989 03:49:14 From: Unknown Subject: Re: NeXT vs Sun3 In article <13246@orstcs.CS.ORST.EDU> kean@nyssa.CS.ORST.EDU.UUCP (Kean Stump) writes: > ...Optical only is a wee bit slow; NeXT has remedied that as of last >Tuesday, by distributing a 40 Mb scsi disk with each optical only cube for the >kernel to swap to (called the "accelerator"). Same price as the original >optical only cube (6500.00 base) and, NeXT is (according to the campus rep) >shipping 40 Mb scsi drives to those who purchased an optical only system. > Very interesting. Can anyone comment on the speed increase with the SCSI?
Date: Sun 21-Oct-1989 08:20:04 From: Unknown Subject: Re: NeXT vs Sun3 > >Can anyone comment on the speed increase with the SCSI? > > I don't have hard data, but from my experience, I'd say the SCSI is 2--3 times faster, and an order of magnitude quiter. But, the floptical was okay on speed and noise. (But it is noisy--when its reading or writeing it sounds like an old daisy wheel printer, or like a modern cash register printing up a receipt.) -Barry Merriman >From: rick@hanauma.stanford.edu (Richard Ottolini)
Date: Sun 21-Oct-1989 17:03:10 From: Unknown Subject: Re: NeXT vs Sun3 The question to answer first with the purchase of any workstation is to what application use you are going to put the machine to. Neither has a particularly large applications base compared to PCs, although Sun has more. If you wish to write programs, particularly graphics, NeXT has better tools. My work requires color, so I have to use Suns now. FORTRAN is easier to implement some types of number crunching, and NeXT doesn't offer it yet. And so on ... evaluate your applications. >From: cs141043@brunix (Ronald Antony)
Date: Sun 21-Oct-1989 20:26:23 From: Unknown Subject: Re: NeXT vs Sun3 There is a third party fortran, as far as I know even with object extensions Ronald . >From: mccalpin@masig3.masig3.ocean.fsu.edu (John D. McCalpin)
Date: Sun 21-Oct-1989 17:24:15 From: Unknown Subject: Re: NeXT vs Sun3 <17434@watdragon.waterloo.edu> <191@med.Stanford.EDU> In article <191@med.Stanford.EDU> rick@hanauma.UUCP (Richard Ottolini) >FORTRAN is easier to implement some types of number crunching, and NeXT >doesn't offer it yet. I'm sure that 1000 people will also post this, but.... Absoft has been shipping FORTRAN for the NeXT for some time. The compiler works pretty well, and the NeXT outperforms the Sun-3 machines at any available clock rate --- provided the Sun-3 does not have the Weitek fpa board! If you want to run numeric-intensive C programs, the NeXT is MUCH faster than a 68881-equipped Sun-3, though the Sun-3 performance can be much enhanced by installing GNU C to replace the standard Sun C compiler. Interestingly enough, the GNU compiler on the NeXT typically runs floating-point intensive codes *faster than FORTRAN*. This is the only machine that I know of that can make that claim. It says very good things about the GNU C optimizer. One could interpret it to say bad things about the Absoft FORTRAN compiler, but I have found that the Absoft compiler actually optimizes pretty well --- almost as well as the Sun-3 f77 compiler with full optimization.
Date: Sun 22-Oct-1989 03:27:56 From: Unknown Subject: Re: NeXT vs Sun3 <17434@watdragon.waterloo.edu> <191@med.Stanford.EDU> <MCCALPIN.89Oct21132415@masig3.masig3.ocean.fsu.edu> (John D. McCalpin) writes: Interestingly enough, the GNU compiler on the NeXT typically runs floating-point intensive codes *faster than FORTRAN*. This is the only machine that I know of that can make that claim. It says very good things about the GNU C optimizer. One could interpret it to say bad things about the Absoft FORTRAN compiler, but I have found that the Absoft compiler actually optimizes pretty well --- almost as well as the Sun-3 f77 compiler with full optimization. Is gcc on the NeXT using inline 68882 instructions? If it is and the Fortran compiler is not, this could explain the speed difference. Saving a function call for every operation will do that... Nick Nick Thompson : nrt@cs.brown.edu | nrt@browncs.bitnet | uunet!brunix!nrt "Lie down on the floor and keep calm." -mantra attributed to John Dillinger >From: jpd00964@uxa.cso.uiuc.edu
Date: Sun 31-Oct-1989 07:07:38 From: Unknown Subject: Re: NeXT vs Sun3 <17434@watdragon.waterloo.edu> <191@med.Stanford.EDU> <MCCALPIN.89Oct21132415@masig3.masig3.ocean.fsu.edu> <NRT.89Oct21232756@planet.cs.brown.edu> In reply to: nrt@cs.brown.edu (Nick Thompson) (Nick Thompson) writes: Is gcc on the NeXT using inline 68882 instructions? If it is and the Fortran compiler is not, this could explain the speed difference. Saving a function call for every operation will do that... I can assure you that the Absoft FORTRAN 77 compiler does generate inline 68882 instructions. There are a slew of optimizations that may be turned on individually and in groups that can vastly effect performance; especially with floating point operations. The Whetstone benchmark almost doubles in timing when using full optimizations. Ken Olson Absoft Corporation < OK TO PORT > This information comes from BIX(R), the BYTE Information Exchange. For additional information about BIX, call 800-227-2983 or 603-924-7681.
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Marcel Waldvogel and Netfuture.ch.