[Info-vax] Unix on A DEC Vax?

Michael Kraemer M.Kraemer at gsi.de
Wed Jan 23 04:07:12 EST 2013


Bill Gunshannon schrieb:
> In article <kdlg5e$fat$1 at solani.org>,
> 	Michael Kraemer <M.Kraemer at gsi.de> writes:
> 
>>Bill Gunshannon schrieb:
>>
>>
>>>Never saw a MIPS that could hold a candle to an Alpha. 
>>
>>Of course it could.
>>Mips-based SGI's made it all through the nineties,
>>and Mips is still alive, whereas Alpha is long dead.
> 
> 
> Not for performance reasons.

Well, I had colleagues who preferred for example
SGI Challenge over contemporary Alpha boxes.
On the workstation market, SGI outsold
Alpha's, iirc.
So the claimed performance lead isn't
evident at all.

> Do you know anything about the demise of Alpha at all?  Development
> stopped over 10 years ago and it is still much in demand. 

Certainly not for performance lead. Just for replacements.
HP stopped selling refurbished boxes more than two years ago.

> And, it
> maintained a performance lead on pretty much everything for at least
> 5 years of no development. 

The last new Alpha's appeared 2003.
I doubt that those machines will beat for example
a Power6 based box of 2008 vintage.
Maybe not even a Power5 or 4 from the years before.
But I might learn something new every day.

> Alpha was killed in order to keep it from
> competing with Itanium because 

Isn't it about time to put this myth to rest?
Maybe onto the same dusty shelf where we finally
have put the "VMS constant" and "VMS rules DoD"?
Alpha was dying long before it was officially EOL'ed
by Compaq. DEC themselves rang the death knell with
that famous intel agreement in 1997.

> it was well known among the engineers
> that Itanium would never win.  And it hasn't.

Those who survive "win".
And the winner is?

> Doesn't require a VAXStation, only an X-terminal.
(snip)
> If only the software that was showing up on PC's actually showed up
> on VMS, but alas....
(snip)
> A VTxxx maybe, but not a VTX2000+.  I know, I got one.  And a couple
> HP X-terminals, too.  Work great on VMS.

I didn't say it was impossible on other systems
nor did I say you needed a VAXstation to do such things.
Just that for a couple of years up to ca 1990 VMS
was very popular in DEC shops especially
on the desktop for typical
technical/scientific purposes, and that almost everybody
preferred his/her own workstation over central big iron.
Exactly the opposite of today's "cloud computing" mantra.

> My experince has, obviously, been differnt.  My VS3100's are fine
> for use at home, but I have a SUN3 that out performs them when it
> comes to number crunching.  OK, an exageration, but my SparcStation
> definitely beats them.  The primary reason I keep them around is to
> run VMS which, the Suns won't run.  :-)

I didn't mention comparison with Unix boxes,
just within the VAX family. For our apps back then,
it turned out that the performance
VS3100 to VAX 6000 to VS3176 was roughly 1:1.5:2.6.
Which led the workstation crowd to the inevitable
question why one should keep a big fat power sucking
6000 when two VS3176 could do the same job for less
money and more desktop capabilities.
A Sun3 is an 68020 machine iirc, so I doubt it will
"beat" a contemporary VS. On par, maybe.
SparcStations OTOH are an entirely new breed,
much younger, so it's no wonder they are faster.




More information about the Info-vax mailing list