les@proper.UUCP (Les Kent) (11/03/85)
Im interested in coprocessor cards for an IBM/AT. I have found four
cards based on the National 320xx. Here are the ones I have found
so far:
o Opus Systems, Cupertino CA
They have two configuraions available. One with the 32032
and the other with the 32016. Memory sizes of 1MB, 2MB and
4MB. It runs System V R2 V2 with virtual memory. They said
they are working on networking support using Excelan software
and hardware but did not give an exact date. They seemed to
be organized and on the ball. The information they sent is
very nicely done and is informative.
o Sritek, Cleveland OH
They have one model available using the 32016 with 2MB of
memory. The information I received says it is expandable
to 8MB. It runs Genix 4.1 BSD with virtual memory support.
I was not very impressed with Sritek for a number of reasons.
First the salesperson the I talked with was not very technical
or knowledgable about the product. Second on the information
I received, dated January 1985, It talks a lot about Genix 4.2
yet when I call, In November 1985, Im told that Genix 4.2 is
not available yet! Most of the other information I received
were reprints from a magazine article in a 1983 Byte and
one from a 1983 PC World. Seems to me like Sritek started
out with some great ideas that died somewhere along the way.
o Definicon Systems, Chatsworth CA
This board was featured in a two part article in the August
and September 1985 Byte. They have one board but allows
options of either a 32032 or a 32016. It comes with as
little as 256K and can be expanded to at least 2MB. This
board also includes two serial ports. I had a very good talk
with a techincal sales engineer and got a lot of in depth
information on their product. He did make a point of saying
that they run there board at 10Mhz and that no other folks
doing coprocessor stuff could do that. At this time they
only have an interface between MS-DOS, and Concurrent DOS.
They have available C, FORTRAN and PASCAL that run stand
alone in the coprocessor card. The person I talked to
said they hope to have UNIX V V2 R2 in about a month. They
same outfit that ported the Opus UNIX is porting their
UNIX so the time frame might be correct.
o Tektronix, Beaverton OR
This board is still a mystery to me. It is a new product
and the field sales offices do not yet know a lot about
it. I do know that it runs Utek, their version of 4.2BSD
that is the same as the 6000 family of workstations. Utek
looks like a very nice version of UNIX that has a mix of
4.2 BSD and System V. They also have there own distributed
file system, networking, improved virtual memory support.
The main thing I don't know right now is if they will be
offering this product as an upgrade product for someone
with an existing IBM/AT or if you have to get the whole
system from them.
These are the boards based on the National 320xx processors. I have
not yet looked at any other types of processors. I would think that
maybe a 68020 would be great but I think it would be hard to fit on
a board that small.
Does anyone have any experience with any of the above boards? Do
you know of any other coprocessor boards? Does XENIX run well
enough on an AT that a coprocessor is not required?
The big problem is that all of the above products have strengths
and weakneses. As an example the Opus product seems to be the
most mature yet does not have networking yet. The Sritek runs
a BSD unix (I prefere) yet they don't have 4.2. The Definicon
product is not complete but seems to have a speed advantage when
it comes out. The Tektronix system seems perfect but might not
be available as an add on, and I have two AT's allready, and also
might be too expensive. I hope to find out more on the Tektronix
product this week.
Les Kent {qantel,dual,amdahl,intelca}!proper!les
Vision Computers dual!proper!les@BERKELEY e-smith@utah-cs.UUCP (Eric L. Smith) (11/04/85)
In article <384@proper.UUCP> les@proper.UUCP (Les Kent) writes: ... >These are the boards based on the National 320xx processors. I have >not yet looked at any other types of processors. I would think that >maybe a 68020 would be great but I think it would be hard to fit on >a board that small. ... I am currently designing an MC68020 based product (not an IBM PC coprocessor, I haven't gone completely insane yet), and based on my knowledge of the 68020 can not see any reason why a person would expect it to occupy more board space than a corresponding 32032 based design. -- ------------------------------------------------------------------------------- Eric L. Smith (801) 581-8100 e-smith@utah-cs.arpa ...decvax!utah-cs!e-smith 3118 Merrill Engineering, University of Utah, Salt Lake City, UT 84112 The opinions expressed herein do not necessarily represent those of the University of Utah, my friends, enemies, computer, or even me. :-) Is this the most magnificant fire you have seen or am I crazy?
baba@spar.UUCP (Baba ROM DOS) (11/04/85)
> o Definicon Systems, Chatsworth CA > > This board was featured in a two part article in the August > and September 1985 Byte... > ...At this time they > only have an interface between MS-DOS, and Concurrent DOS. > They have available C, FORTRAN and PASCAL that run stand > alone in the coprocessor card. The person I talked to > said they hope to have UNIX V V2 R2 in about a month. The > same outfit that ported the Opus UNIX is porting their > UNIX so the time frame might be correct. Odd. The Opus people I've met gave me the distinct impression that they had done their own port. Who are these people Definicon is speaking of? ROM DOS
phil@osiris.UUCP (Philip Kos) (11/08/85)
A board featuring a 32032 with 2Mbytes of primary memory running SVR2 is considered to be a *coprocessor* in a PC/AT? Forgive the hyperbole, but isn't that something like towing a Peterbilt behind a Pinto? (:-) What ever happened to truth in advertising (:-)^n)??? Perhaps we'll see ads for "PC/AT coprocessor enclosures for 32032 boards" real soon now. Phil Kos The Johns Hopkins Hospital Baltimore, MD "Mona, you're a brick!" ...!gamma!cp1 \ - C. Savage, Jr. ...!decvax!decuac \ ...!unc!brl-bmd /\ ...!allegra!umcp-cs / \ !aplvax!osiris!phil
rxb@rayssd.UUCP (richard a brooks) (11/14/85)
=== REFERENCED ARTICLE =================================== > Relay-Version: version B 2.10 5/3/83; site rayssd.UUCP > Posting-Version: version B 2.10.1 6/24/83; site utah-cs.UUCP > Path: rayssd!raybed2!linus!philabs!cmcl2!seismo!utah-cs!e-smith > From: e-smith@utah-cs.UUCP (Eric L. Smith) > Newsgroups: net.micro.16k,net.micro.pc > Subject: Re: 32032 coprocessor board for IBM/AT > Message-ID: <3522@utah-cs.UUCP> > Date: Sun, 3-Nov-85 21:10:28 EST > Article-I.D.: utah-cs.3522 > Posted: Sun Nov 3 21:10:28 1985 > Date-Received: Tue, 5-Nov-85 21:48:31 EST > References: <384@proper.UUCP> > Reply-To: e-smith@utah-cs.UUCP (Eric L. Smith) > Organization: University of Utah Ada-to-Silicon project > Lines: 21 > Xref: rayssd net.micro.16k:305 net.micro.pc:4204 > > In article <384@proper.UUCP> les@proper.UUCP (Les Kent) writes: > ... > >These are the boards based on the National 320xx processors. I have > >not yet looked at any other types of processors. I would think that > >maybe a 68020 would be great but I think it would be hard to fit on > >a board that small. > ... > > I am currently designing an MC68020 based product (not an IBM PC > coprocessor, I haven't gone completely insane yet), and based on my > knowledge of the 68020 can not see any reason why a person would expect > it to occupy more board space than a corresponding 32032 based design. > -- > ------------------------------------------------------------------------------- > Eric L. Smith (801) 581-8100 e-smith@utah-cs.arpa ...decvax!utah-cs!e-smith > 3118 Merrill Engineering, University of Utah, Salt Lake City, UT 84112 > > The opinions expressed herein do not necessarily represent those of the > University of Utah, my friends, enemies, computer, or even me. :-) > > Is this the most magnificant fire you have seen or am I crazy? ========================================================== *** REPLACE THIS LINE WITH YOUR MESSAGE ***