[comp.sys.ibm.pc] 486/25 Power Platform for PS/2 70-A21

ken@gatech.edu (Ken Seefried III) (07/12/89)

I have a little bit of marketing info on this product, and I have a
few questions:

What will it cost and when will it ship?

Will AIX run on it (it should)?

What kind of real-world performace will you gain from a 486/25 over a
386/25 (marketing rubbish claims 80% on 'buisiness applications' 
(whatever they are) and 4 times better on 'compute intensive applications' 
(whatever they are))

If it delivers what they claim, might actually be a reason to buy a
PS/2-70 (heaven forbid...;').

Thanks...

	...ken seefried iii
	   ken@gatech.edu

mm42@prism.gatech.EDU (Mark A. Mitchell GTRI/STL/MSD) (07/12/89)

In article <19000@gatech.edu> ken@gatech.edu (Ken Seefried III) writes:
>I have a little bit of marketing info on this product, and I have a
>few questions:
>
>What will it cost and when will it ship?
>
According to PC Week, $3995 and by the end of the year.

>What kind of real-world performace will you gain from a 486/25 over a
>386/25 (marketing rubbish claims 80% on 'buisiness applications' 
>(whatever they are) and 4 times better on 'compute intensive applications' 
>(whatever they are))
>
PC Week's tests of the "Power Platform" (that's what IBM calls it) included
a CPU sort and an unspecified Floating Point test.  The time in seconds
for three machines:
				CPU sort		Floating Point
	IBM Model 70 A21       	6.1			5.2
	Compaq 386/33 MHz	4.7			3.5
	IBM 486 PP		3.2			2.3

So it looks about 2 times as fast.  

>If it delivers what they claim, might actually be a reason to buy a
>PS/2-70 (heaven forbid...;').
>

-- 
Mark A. Mitchell  GTRI/STL/MSD
Georgia Institute of Technology, Atlanta Georgia, 30332
uucp:	  ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!mm42
Internet: mm42@prism.gatech.edu

rkh@mtune.ATT.COM (Robert Halloran) (07/12/89)

In article <991@hydra.gatech.EDU> mm42@prism.gatech.EDU (Mark A. Mitchell  GTRI/STL/MSD) writes:
>In article <19000@gatech.edu> ken@gatech.edu (Ken Seefried III) writes:
>>I have a little bit of marketing info on this product, and I have a
>>few questions:
>>
>>What will it cost and when will it ship?
>>
>According to PC Week, $3995 and by the end of the year.
>>If it delivers what they claim, might actually be a reason to buy a
>>PS/2-70 (heaven forbid...;').

As I understand it, this is a 486 replacement board for an existing PS/2-70.
Between buying the initial PS/2 and this board, you're up to around $10-11K,
right?  As it is safe to assume that Intel isn't restricting chip supply
to IBM, what does this buy you over waiting for some probably-very-small
period until Compaq/Zenith/ATT/.........  get their machines out, unless
you already own the PS/2?  A perplexed mind wants to know....

						Bob Halloran
=========================================================================
UUCP: att!mtune!rkh				Internet: rkh@mtune.ATT.COM
USPS: 17 Lakeland Dr, Port Monmouth NJ 07758	DDD: 201-495-6621 eve ET
Disclaimer: If you think AT&T would have ME as a spokesman, you're crazed.
Quote: "Where does he get those wonderful toys?" - Joker, BATMAN

ken@capone.gatech.edu (Ken Seefried iii) (07/13/89)

In article <8032@mtune.ATT.COM> rkh@mtune.UUCP (Robert Halloran) writes:
>
>As I understand it, this is a 486 replacement board for an existing PS/2-70.
>Between buying the initial PS/2 and this board, you're up to around $10-11K,
>right?  

Correct on all counts...

>        As it is safe to assume that Intel isn't restricting chip supply
>to IBM, what does this buy you over waiting for some probably-very-small
>period until Compaq/Zenith/ATT/.........  get their machines out, unless
>you already own the PS/2?  A perplexed mind wants to know....
>
>						Bob Halloran

Well...lets see if I can name a few without being labeled an IBM
flag-waver...

I can get both the PS/2-70 A21 and the 486/25 at substatial discount
as a student.  Price becomes less of an issue...

Speaking of price...have you recently priced a 25Mhz 386 with 2MB RAM
and a 120MB hard disk from Compaq?  Certainly not much cheaper than the
PS/2 equivalent...I don't expect their 486 boxes to buck this trend.
Indeed, most things that I have read peg mainstream 486 boxes in the
$15K-$20K ballpark, at least for some time (feel free to correct me if
wrong)...

The 486/25 availible real soon now.  Not nessesarily so with the
Compaq/Zenith/ATT hoard...

Perhaps I want or need MCA (and there are some reasons why I might),
and ATT/Compaq/Zenith don't have it...

Same as above for AIX...

Now, don't get me wrong...I'm no IBM fan, and I'm certainly no Intel
fan, but the PS/2-70 is a slick product, and the 486 upgrade is a very
slick product (I can hear the salesman now: "buy the 70-A21 now, and
you'll be able to upgrade it to a 486 when you need it, by just poping
in this little board!  No slots taken!"  Great stuff for marketing).
I have to give credit where it is due...*sigh*

	...ken seefried iii
	   ken@gatech.edu
	ken seefried iii	...!{akgua, allegra, amd, harpo, hplabs, 
	ken@gatech.edu		masscomp, rlgvax, sb1, uf-cgrl, unmvax,
	                      ut-ngp, ut-sally}!gatech!ken

rogers@falcon.SRC.Honeywell.COM (Brynn Rogers) (07/13/89)

In article <8032@mtune.ATT.COM> rkh@mtune.UUCP (Robert Halloran) writes:
<In article <991@hydra.gatech.EDU> mm42@prism.gatech.EDU 
                        (Mark A. Mitchell  GTRI/STL/MSD) writes:
<>In article <19000@gatech.edu> ken@gatech.edu (Ken Seefried III) writes:
<>>I have a little bit of marketing info on this product, and I have a
<>>few questions:
<>>
<>>What will it cost and when will it ship?
<>>
<>According to PC Week, $3995 and by the end of the year.
<>>If it delivers what they claim, might actually be a reason to buy a
<>>PS/2-70 (heaven forbid...;').

<As I understand it, this is a 486 replacement board for an existing PS/2-70.
<Between buying the initial PS/2 and this board, you're up to around $10-11K,
<right?  As it is safe to assume that Intel isn't restricting chip supply
<to IBM, what does this buy you over waiting for some probably-very-small
<period until Compaq/Zenith/ATT/.........  get their machines out, unless
<you already own the PS/2?  A perplexed mind wants to know....

I think Info week said that the PS/2-70A was $12K to begin with.  also
remember that that the 70A was recalled and delayed for months to fix
a small problem [It didn't work or something].

The 486 would like to read 64 bit words from memory if it could, and
It can read something like 16 Bytes in 4 or 5 clock cycles in burst mode.
with 32 bits read every 2 or 3 clocks like a 386 does, you can see that
this thing will really smoke once it gets a system built around it.

 Brynn Rogers    Honeywell S&RC        rogers@src.honeywell.com
       612-782-7737                use this address if your reply bounces

mm42@prism.gatech.EDU (Mark A. Mitchell GTRI/STL/MSD) (07/13/89)

In article <25366@srcsip.UUCP> rogers@falcon.UUCP (Brynn Rogers) writes:
>I think Info week said that the PS/2-70A was $12K to begin with.

Well, it was, but IBM is dropping the price to $8995 to encourage folks
to buy one and wait for the 486 board.
-- 
Mark A. Mitchell  GTRI/STL/MSD
Georgia Institute of Technology, Atlanta Georgia, 30332
uucp:	  ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!mm42
Internet: mm42@prism.gatech.edu

jwi@lzfme.att.com (Jim Winer @ AT&T, Middletown, NJ) (07/13/89)

> Ken Seefried III writes:
>
> What kind of real-world performace will you gain from a 486/25 over a
> 386/25 (marketing rubbish claims 80% on 'buisiness applications' 
> (whatever they are) and 4 times better on 'compute intensive applications' 
> (whatever they are))
> 
> If it delivers what they claim, might actually be a reason to buy a
> PS/2-70 (heaven forbid...;').

What you will have when you plug in this board is a 486 with
built-in cache running on top of a second cache designed for a
386. A secondary cache designed for the 486 should have considerably
different charactistics than result from using a primary 386 cache
as a secondary 486 cache.

You also get a built-in FP co-processor -- which may or may not
conflict with your existing 80387, but which will certainly obsolete
it.

So: 1 80386-25 and 1 80387-25 down the drain = $1000. Replacement
486 board at least $1500. Why not just wait and get a machine that's
designed for the 486?

More importantly, several other plug-in 486 replacement boards will
be annouunced shortly -- don't get stuck with the MCA and
IBM. Almost everybody else (in the same price range) does it better.
For the samme amount of money in 6 months you can buy a real 486 at
33 or 38MH from DELL or Compaque and have something that's well
designed. For less money, you will be able to buy one from CompuAdd
that's almost as good and still outperforms the IBM.

Jim Winer ..!lzfme!jwi (Usually unable to reply to email outside AT&T)

Those persons who advocate censorship offend my religion.

Upuaut:	a wolf-headed Egyptian deity | Voodoo: the art of sticking ideas
	assigned as Guidance System  |         into people and watching
	for the Barque of Ra.        |         them bleed.

The opinions expressed here are not necessarily  

tim@cit-vax.Caltech.Edu (Timothy L. Kay) (07/13/89)

In article <19000@gatech.edu> ken@gatech.edu (Ken Seefried III) writes:

>What will it cost and when will it ship?

The upgrade board costs $3,999, and you need the Model 70 A21 to plug
it into.  That costs about $8,995.  It won't ship until Intel delivers
production quantities.

>Will AIX run on it (it should)?

Word has it that it won't run AIX nor OS/2 just yet.  I am not sure
if this is do to bugs in early silicon or just plain incompatibilities.
I would guess its the bugs.

>What kind of real-world performace will you gain from a 486/25 over a
>386/25 (marketing rubbish claims 80% on 'buisiness applications' 
>(whatever they are) and 4 times better on 'compute intensive applications' 
>(whatever they are))

Yeah, maybe a factor of two.

>If it delivers what they claim, might actually be a reason to buy a
>PS/2-70 (heaven forbid...;').

Or, you could just wait and buy an AST or a Compaq, which will most
certainly come out at about the same time, cost less, and perform
better.  And, it will have an ISA bus, so you can actually find things
to plug in.

People should realize by now that IBM hasn't had a personal computer
that is faster than the competition since Compaq opened its doors.

Tim

werner@aecom.yu.edu (Craig Werner) (07/13/89)

In article <25366@srcsip.UUCP>, rogers@falcon.SRC.Honeywell.COM (Brynn Rogers) writes:
> 
> <As I understand it, this is a 486 replacement board for an existing PS/2-70.
> <Between buying the initial PS/2 and this board, you're up to around $10-11K,
> <right?  

	I've actually seen a model 70 running the 486, side by side with one
not running one.  In the benchmarks I saw, there were approximately three
dots on the '486 machine for every one on the '386 machine.  It was actually
slightly more than 3:1. And the '486 machine had an extra window open (the IBM
rep offered to close it, but I told him not to bother), so that brings the
performance probably as high as 3.5:1 But remember, that's 25 vs. 25.  If
one starts having 33MHz '386s, the gap narrows, but so does the cost. Me
personally, a 16MHz 80386 still seems blinding. I'm only months beyond 4.77!
-- 
	        Craig Werner   (future MD/PhD, 4.5 years down, 2.5 to go)
	     werner@aecom.YU.EDU -- Albert Einstein College of Medicine
              (1935-14E Eastchester Rd., Bronx NY 10461, 212-931-2517)
               "Why is it that half the calories is twice the price?"

rogers@falcon.SRC.Honeywell.COM (Brynn Rogers) (07/13/89)

In article <1485@lzfme.att.com> jwi@lzfme.att.com (Jim Winer @ AT&T, Middletown, NJ) writes:
<So: 1 80386-25 and 1 80387-25 down the drain = $1000. Replacement
<486 board at least $1500. Why not just wait and get a machine that's
<designed for the 486?
^^^^   Just for the record, IBM charges $1295 for a 387-25 and 
 approx. $2000 for 80386/80387 CPU board for one of these. 
 AND They have to have trade in on it! (so your out a extra $2k for
 your old CPU board when they put a power platform in)

Read the last few INFO WEEK newspapers for useful info. (all have articles
about the power platform)

<More importantly, several other plug-in 486 replacement boards will
<be annouunced shortly -- don't get stuck with the MCA and
<IBM. Almost everybody else (in the same price range) does it better.
<For the samme amount of money in 6 months you can buy a real 486 at
<33 or 38MH from DELL or Compaque and have something that's well
<designed. For less money, you will be able to buy one from CompuAdd
<that's almost as good and still outperforms the IBM.

I agree with everything but the last line. anything thousands of dollars less 
that still outperforms IBM is BETTER than IBM! (not almost as good)

 Brynn Rogers    Honeywell S&RC        rogers@src.honeywell.com
       612-782-7737                use this address if your reply bounces

wtm@neoucom.UUCP (Bill Mayhew) (07/14/89)

What IBM [probably] forgot to show you was a real application that
used a lot of I/O: something like a Microsoft C5.0 compile or
something that did a lot of output to the screen.

I have a feeling that timing a C compile with a stopwatch would
yield speed-up of about 1.5 on the PS2/70 with the 486 daughter
board.

Mind you, that isn't exactly chopped liver, but that is pretty
expensive with the projected list price of slightly less than $4K
(US) for the upgrade.  These days, you can get an entire PS/2 model
80-171 (16MHz) system (if you can find a dealer with one in stock)
for about that much.  The price of the upgrade is somewhat
ameliorated by the fact that the 486 contains the equivalent of the
80387 FPU; this effectively discounts the cost of the upgrade by
$500, assuming that you'd like to have the FPU chip in your
system.

Curious that the 486 board came out for the model 70, rather than
the model 80.  So much for the 'flagship' of the PS/2 line, I
guess.  (Somewhat nonplussed, as I have a model 80).

Bill
wtm@impulse.UUCP   or   wtm@neoucom.UUCP

chasm@attctc.DALLAS.TX.US (Charles Marslett) (07/16/89)

In article <8032@mtune.ATT.COM>, rkh@mtune.ATT.COM (Robert Halloran) writes:
> As I understand it, this is a 486 replacement board for an existing PS/2-70.
> Between buying the initial PS/2 and this board, you're up to around $10-11K,
> right?  As it is safe to assume that Intel isn't restricting chip supply
> to IBM, what does this buy you over waiting for some probably-very-small
> period until Compaq/Zenith/ATT/.........  get their machines out, unless
> you already own the PS/2?  A perplexed mind wants to know....

Its cheaper than the $15,000 Compaq box, just about as fast, and probably
breaks less software...

Zenith's list price will be higher than Compaq's and you won't be able to
get it for a year -- but the street price will be reasonable.

AT&T's will be built by Olivetti (since Intel won't be shipping boards yet
-- can't figure out the CPU chip timing goofs).  That's enough to sell
Model 70s alone.

[P.S.  lots of smileys....]

> 						Bob Halloran

Charles Marslett
chasm@attctc.dallas.tx.us

davidsen@sungod.crd.ge.com (William Davidsen) (07/22/89)

In article <19000@gatech.edu> ken@gatech.edu (Ken Seefried III) writes:
| I have a little bit of marketing info on this product, and I have a
| few questions:
| 
| What will it cost and when will it ship?

  I saw the product anouncement on this last week, and I believe it
ships 4th quarter. The price is breathtaking... $4k, and they *keep*
your 386 board! There was a sentence saying something like ``the
customer may retain his 387 chip if for some reason he wants to.''

  For six grand I bet I can buy a whole computer by 2Q90, providing I
get it from someone else. Watch _Computer Shopper_ on this one. The IBM
upgrade is good for those with lots of money and a desperate need, in my opinion.

	bill davidsen		(davidsen@crdos1.crd.GE.COM)
  {uunet | philabs}!crdgw1!crdos1!davidsen
"Stupidity, like virtue, is its own reward" -me

richard@uhccux.uhcc.hawaii.edu (Richard Foulk) (07/31/89)

} I can get both the PS/2-70 A21 and the 486/25 at substatial discount
} as a student.  Price becomes less of an issue...

Their prices are so high that even a "substantial" discount doesn't
cut it.  Check into a decent clone, of which there are many.

Compaq doesn't count.  They're way over-priced too.

} The 486/25 availible real soon now.  Not nessesarily so with the
} Compaq/Zenith/ATT hoard...

IBM simply invoked it's usual sleezy approach to marketing and pre-
announced.  Everyone else will sell one, cheaper, faster and better.

How many times do you have to see the cycle to believe it?

} Perhaps I want or need MCA (and there are some reasons why I might),
} and ATT/Compaq/Zenith don't have it...

If you've already bought into MCA then you're stuck.

If more people see through the pre-announcements and other endless
sleeze, and just say no, then maybe IBMs attempt to bring the world
back to propietary hardware standards (MCA) will die away.


Richard

davidsen@sungod.crd.ge.com (William Davidsen) (08/01/89)

In article <4466@uhccux.uhcc.hawaii.edu> richard@uhccux.UUCP (Richard Foulk) writes:

| IBM simply invoked it's usual sleezy approach to marketing and pre-
| announced.  Everyone else will sell one, cheaper, faster and better.
	Probably sooner, too.
| 
| How many times do you have to see the cycle to believe it?

  Actually IBM is in a no win situation here... when they didn't
pre-announce things they got a lot of noise because they "didn't let the
public in on their future products for planning purposes." Now they get
noise for pre-anouncing to spoil other's market.

  I'm certainly not an IBM lover, but in this case I think they are
being no more disreputable than any other vendor.

  I expect to see a comparable machine for less money than the list
price of the upgrade, and delivering before IBM does. The list price is
$4k for the 486, and you have to give IBM your 386, worth about $1500 at
their prices. If you have an 80387 add $1200 their (new, lower) price.

  Question: the IBM product anouncement states that you *must* hand in
your 386 when they upgrade your machine. What is IBM going to do with
all those boards? Anyone from IBM want to comment?

	bill davidsen		(davidsen@crdos1.crd.GE.COM)
  {uunet | philabs}!crdgw1!crdos1!davidsen
"Stupidity, like virtue, is its own reward" -me