[comp.sys.ibm.pc.hardware] EISA -- Is it worth it?

gurganus@palomino.ecn.purdue.edu (James P Gurganus) (05/31/91)

I posted this before, but it never appeared, so my apologies if this article
has appeared twice...

I'm REALLY close to buying an 486-33 EISA system.  I found one with SVGA and
everything for right around $4000.  After calling 40 dealers or so, I've found
that the majority of them think that EISA isn't worth it since little or 
nothing is available for it and that nobody will create any hardware for it
as long as I live, etc., etc.  Then I read popular magazine articles saying
"unleash the power of your 486 with EISA."  I'm slightly confused.

I know the difference between ISA and EISA and that if I don't have any 32-bit
cards, it won't make a whole lot of difference.  (Or so I'm told.)  But I'd
kinda like to be prepared whatever happens.

What's the general net-feeling on EISA?  Is there any hardware available or
in development for it?  I heard somewhere that there was a 32-bit SCSI 
controller out.  

Even though this price is pretty good, I know I can get a fast system from
ARES for $300 less or so.  ARES doesn't have any EISA systems out.

Thanks for any info or advice....

James Gurganus
(gurganus@ecn.purdue.edu)

solomon@chaos.utexas.edu (Thomas Solomon) (06/01/91)

In article <gurganus.675629651@palomino.ecn.purdue.edu> gurganus@palomino.ecn.purdue.edu (James P Gurganus) writes:
>
>everything for right around $4000.  After calling 40 dealers or so, I've found
>that the majority of them think that EISA isn't worth it since little or 
>nothing is available for it and that nobody will create any hardware for it
>as long as I live, etc., etc.  Then I read popular magazine articles saying
>"unleash the power of your 486 with EISA."  I'm slightly confused.

This business about nothing being available and "that nobody will create
any hardware for it as long as [we] live" is just not true.  Our lab is
planning on buying an EISA machine for data taking and analysis.  The
machine that we are buying will have an EISA disk controller (worth the
value of the EISA bus alone) and an image processing system that uses
the EISA interface for 32-bit data transfers.  In addition, there are
EISA ethernet cards currently available, along with data acquisition
hardware (from National Instruments) that takes advantage of the EISA
to get higher throughput.  These are all things that are available _now_.
UNDOUBTABLY, companies will make more EISA products in the future, once
enough EISA machines are out there and once people start realizing the
full potential of EISA.  With 32 bits versus 16, and 33 MHz capabilities
versus 8 MHz, an EISA system can transfer data to or from peripherals
up to 8 times faster than ISA.  I think that it is inconceivable that
(in this highly competitive computer market) companies will not try to
take advantage of that power.

Granted, our applications require more bus power than many applications,
since we are interfacing our computer to experiments.  But I am still
of the opinion that it is worth a few extra hundred dollars to buy a
system that will be able to keep up with the times.  So if you think
that you may need to do some interfacing, or if you might need a super-
fast disk controller in the future, then go with the EISA.  On the other
hand, if you are planning on using your machine for word processing,
data base stuff, spreadsheets or the kind of stuff that most businesses
buy PCs for (i.e., tasks in which the computer need not communicate
very much with peripherals), then you will never need the EISA power.

                                      Tom Solomon
                                      solomon@chaos.utexas.edu