haral@bnl.UUCP (Haral Tsitsivas) (04/06/84)
Now that the VAX 11/785 has been announced (this week's CAEM Expo, at the Los Angeles Airport Hilton) does anyone know of any work done to get UNIX running on the 785? DEC offers an upgrade package which would be good to have (50% performance improvement) if your machines run with a load factor average of 20 to 25.
gwyn@brl-vgr.ARPA (Doug Gwyn ) (04/08/84)
About the only reason I can see to fork over $91K ($95K with FPA) for an upgrade of your VAX-11/780 to a VAX-11/785 is that you need more CPU power ON THE VAX YOU HAVE. If you can acquire a second computer system, there are many more cost-effective ways to get about 50% of a VAX-11/780; for example, you could buy 30 Macintoshes and get 1500% of the power of a 780 (not the best alternative, but it is a remarkable contrast).
rcd@opus.UUCP (04/09/84)
<> >...If you can acquire a second computer >system, there are many more cost-effective ways to get about 50% of >a VAX-11/780; for example, you could buy 30 Macintoshes and get 1500% >of the power of a 780 (not the best alternative, but it is a remarkable >contrast). Yes, and if you acquire multiple machines to handle one computing load, you get an added bonus - EDUCATION, as follows: You can find out how networks work, and how they fail. You can discover that you have to balance your user load across machines, and that the load will invariably change faster than you can adapt. When you start to do file backup, you'll discover that either you need tape drives on every machine or you get to learn about file system backup across the network. And if users don't have most of the files they need for their work on their home machine, the network overhead will eat you alive. (I think I want a statement of Grosch's law here??? My memory fails me.) I'm not flaming about the idea of "a computer on every desktop" - I think that's a good idea - but WATCH OUT. If you've got lots of little jobs going on, fine - but if you really have a community of people trying to work on one large project, you may be going the wrong way. (Using nine little computers to do the job of one big one can be like trying to have nine women produce a baby in one month.) -- "A friend of the devil is a friend of mine." Dick Dunn {hao,ucbvax,allegra}!nbires!rcd (303) 444-5710 x3086
darrelj@sdcrdcf.UUCP (Darrel VanBuer) (04/10/84)
The biggest problem I have with large systems and even larger systems like a
785 is uptime reliability. Our 780 occupies some eight cabinets (processor,
extra massbus and unibus space, disks, tapes) most of which contain
components whose failure kill the whole system. To support the same 50 users
with workstations would be about 50 quarter-size cabinets (more total
hardware) but none of whose failure affects more than a few users.
I grant that our VAX has excellent availability on the whole, but it's size
makes problems which do occur time-consuming to isolate and repair, and
require much more scheduled down-time for enhancements and maintenance.
A further benefit of workstations is the availability of a lot of spare
cycles which can be devoted to a more supportive user interface for
interactive activities (e.g. large bitmap display per user vs. a limit of
six VAX stations per 780).
--
Darrel J. Van Buer, PhD
System Development Corp.
2500 Colorado Ave
Santa Monica, CA 90406
(213)820-4111 x5449
...{allegra,burdvax,cbosgd,hplabs,ihnp4,sdccsu3,trw-unix}!sdcrdcf!darrelj
VANBUER@USC-ECL.ARPAchris@umcp-cs.UUCP (04/11/84)
There is also this to consider: how hard is it to convince the
management (whatever - here we have the Maryland Information Systems
Divistion to bog us down) to allow you to buy more computers? And
how hard is it to convince them to allow you to upgrade existing
computers? Guess which wins! The least cost effective, usually.
*Sigh*
--
In-Real-Life: Chris Torek, Univ of MD Comp Sci
UUCP: {seismo,allegra,brl-bmd}!umcp-cs!chris
CSNet: chris@umcp-cs ARPA: chris.umcp-cs@CSNet-Relayrpw3@fortune.UUCP (04/21/84)
#R:bnl:-40600:fortune:11600090:000:970 fortune!rpw3 Apr 20 18:41:00 1984 +-------------------- | I agree with the sentiments, but not the arithmetic. 30 Macs == 15 780's | implies that 1 Mac = 1/2 780. I think 30 Macs = 2 780's would be a | LOT closer. | <mike +-------------------- Uh... depends on what KIND of cpu ticks you need for your job. IF 1. You can fit it memory-resident into a Mac, 2. It is not floating-point or even multiply/divide intensive, 3. It involves a lot of pointer arithmetic, character banging, and terminal I/O (like a typical editor), 4. You don't need fast file I/O (or your Mac already has a "net disk"), THEN a Mac is indeed almost exactly 40% of a VAX-11/780. If the app violates #1, YOU LOSE. If it violates #2 or #4, your figure of 1/15'th of a VAX may be high. You have to compare apples and... Oh, never mind. Rob Warnock UUCP: {ihnp4,ucbvax!amd70,hpda,harpo,sri-unix,allegra}!fortune!rpw3 DDD: (415)595-8444 USPS: Fortune Systems Corp, 101 Twin Dolphin Drive, Redwood City, CA 94065
gwyn@BRL-VLD.ARPA (04/25/84)
From: Doug Gwyn (VLD/VMB) <gwyn@BRL-VLD.ARPA> I was under the impression that a 68010 at 12MHz was about 1/2 a VAX-11/780. If this isn't true then I stand corrected.. In any case, my point is that the 11/785 does not appear to be cost- effective in today's market. This is my own opinion, not necessarily that of my employer.
mwm@ea.UUCP (04/29/84)
#R:bnl:-40600:ea:13500010:000:674 ea!mwm Apr 15 23:10:00 1984 /***** ea:net.unix-wizar / brl-vgr!gwyn / 6:19 pm Apr 8, 1984 */ About the only reason I can see to fork over $91K ($95K with FPA) for an upgrade of your VAX-11/780 to a VAX-11/785 is that you need more CPU power ON THE VAX YOU HAVE. If you can acquire a second computer system, there are many more cost-effective ways to get about 50% of a VAX-11/780; for example, you could buy 30 Macintoshes and get 1500% of the power of a 780 (not the best alternative, but it is a remarkable contrast). /* ---------- */ I agree with the sentiments, but not the arithmetic. 30 Macs == 15 780's implies that 1 Mac = 1/2 780. I think 30 Macs = 2 780's would be a LOT closer. <mike