🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.sys.ibm.pc.misc
Article #9168

Re: EISA vs. ...

#9168
From: Harold G. Andrews II
Date: Fri, 12 Apr 1991 18:43
70 lines
4548 bytes
In article <91101.194645U35832@uicvm.uic.edu> U35832@uicvm.uic.edu writes:
>I understand there is the EISA architecture and the regular arch.
>For 486's what would be the best to consider?     
>
That really depends on what you are doing and how much more you are willing
to pay for EISA boards vs ISA boards.  Right now, the availibility of EISA
boards is limited, and those that are available are more than just slightly
expensive.  For instance, Mylex sells three EISA boards (not including their
EISA mother boards) that, performance-wise, do things no ISA board can hope
to accomplish in the near future.  They are the only company I am aware of
currently marketing EISA boards.

>Which is the best architecture?
>
This rather depends on you definition of "best".  If flexibity is your thing
then the EISA architecture is probably the way to go.  You may want to do
some kind of cost-benefit analysis to see if you really need the increased
performance and flexibility the EISA architecture offers over the older ISA
architecture.

>What the hell is the difference?
>
The history as I understand it:  When IBM first came out with their PC way
back when (circa 1981 I believe) they had an 8-bit bus.  When they decided
to come out with an 80286 based machine (the IBM PC/AT) they decided to upgrade
the bus to accept 16 bit cards as well as the older 8 bit cards.  Thus was
born ISA (the Industry Standard Architecture).  This downward compatibility was
one of IBM's major selling points for the AT's (this did not come without its 
problems, however).  By this time, though IBM had some competition.  Compaq 
and others were coming out with lower priced machines with better performance, 
more features etc.  IBM then figured it had had enough of its competitors 
meddling in the market, and decided to throw a monkey wrench into the market.  
The impression I get is that upper management at IBM felt that they had been
successful thus far in directing where the market went.  After all, their
8086 machines had become an instant standard.  IBM designed a new architecture,
called the micro-channel architecture (or MCA) which had a 32 bit bus.  The 
catch was that this new architecture was not downwardly compatible with the 
existing 16-bit ISA.  In response to this, a consortium of several companies 
(Compaq, Hewlett-Packard, Tandy and four others that I forget) got together to 
design a 32 bit downwardly compatible bus architecture which has become known 
as EISA (the extended industry standard architecture).  Now, EISA will allow 
you to use the older 16 bit cards, the even older 8 bit cards (if you can find 
them any more) and the new 32 EISA cards.  You probably won't notice any 
improvement in performance using the 16 bit cards in an EISA slot (though I 
could be mistaken).  You will, however, see a sizable increase in the speed 
with EISA cards in EISA slots (as opposed to an ISA card in an EISA slot
with the same "purpose in life" as the EISA card in question).

>
>Shawn
>University of Illinois at Chicago
>
>
>P.S. Which one has the best future? (Which one will outlast the other,
>in the long run?)                                            
>
I don't know.  I can't see into the future.  At a guess, I would believe
that the EISA bus, being the next logical progression in this bus architecture
evolution, would probably outlive the older ISA architecture.  Others will
disagree.  It's a matter of who you ask.

-Andy

*******************************************************************************
* Harold G. "Andy" Andrews II             *  "Many the man whose punctuality  *
* andrewsh@lonex.radc.af.mil              *   serves only to warm his chair." *
* Rome Laboratory/IRRE                    *                                   *
* Griffiss AFB, NY 13441-5700             *   - M. Kabrisky                   *
* (315) 330-7788  (DSN Prfx 587)          * (Not an official  USAF viewpoint) *
*******************************************************************************
Technical Details
Message-ID: <1991Apr12.184341.24843@lonex.radc.af.mil>
Path: arc1-ca.pugleaf.net!utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!mahendo!wlbr!lonex.radc.af.mil!andrewsh
References: <91101.194645U35832@uicvm.uic.edu>

Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!mahendo!wlbr!lonex.radc.af.mil!andrewsh
From: andrewsh@lonex.radc.af.mil (Harold G. Andrews II)
Newsgroups: comp.sys.ibm.pc.misc
Subject: Re: EISA vs. ...
Message-ID: <1991Apr12.184341.24843@lonex.radc.af.mil>
Date: 12 Apr 91 18:43:41 GMT
References: <91101.194645U35832@uicvm.uic.edu>
Organization: Rome Laboratory, Griffiss AFB, NY USA
Lines: 70