[Info-vax] Vaxes shutting off this week

[Bill Gunshannon]

In article <goevqd$a2o$2 at pcls4.std.com>,
	moroney at world.std.spaamtrap.com (Michael Moroney) writes:
> billg999 at cs.uofs.edu (Bill Gunshannon) writes:
> 
>>In article <goeb8m$fr6$2 at pcls6.std.com>,
>>	moroney at world.std.spaamtrap.com (Michael Moroney) writes:
>>> 
>>>>   Hmm.. I wonder what would have happened had VAX continued on the path 
>>>>   and then did what all of the other processors had to do get more 
>>>>   through-put, go multi-core with more front-end cache...
>>> 
>>>>   Hmm.. I wonder what would have happened had PDP-11 continued on the path 
>>>>   and then did what all of the other processors had to do get more 
>>>>   through-put, go multi-core with more front-end cache...
>>> 
>>> A big difference is that the PDP and the VAX ran their courses and were
>>> discontinued at an appropiate time.  
> 
>>Some people might not agree with that notion.  PDP-11 development was
>>continued long after Dec sold it off to Mentec.  And I mean processor
>>development, not just OS and software support.  And many people here
>>have suggested that while Alpha was a great idea there was a place for
>>VAX to continue as well.  The world never went completely RISC and the
>>VAX was probably better than any other non-RISC processor.
> 
> Somewhat supports my claim.  The PDP-11 was allowed to make money for
> pretty much as long as it was able to, for somebody.  

And longer except the current owners have decided not to deal with it
anymore.  Much to the lament of more people than I can count who are
still using PDP-11's everyday fro real work.

>                                                       But being restricted
> to 64K of memory at once is a big restriction, except for smaller control
> modules. 

Hmmm...  Let's see.  As recently as about 10 years ago I went out to
Joisey to pick up a bunch of RA disks.  The place I went to had converted
all their data storage over to SCSI. They were, however, still happily
running all of their business on PDP-11's.  They were (are?) one of the
largest mortgage brokerage firms on the east coast.  Doing it all on RSTS.
There are a lot of things other than "smaller control modules" that can
still be done quite effectively on less hardware than a Cray.

>            (how much memory is really needed for a computer-controlled 
> traffic light even with lots of sensors to operate?  How about your
> microwave?)  

Well, most of my PDP-11's actually have 2-4 Meg of memory and handle it
within the 64K space quite well.  Even UCSD-Pascal could do "virtual
arrays" making them limited by disk space rather than memory size.  Also,
with overlays program size isn't much of a problem either.  I may not
be able to render the graphics for "Jurasic Park" on my PDP-11, but there
are a lot of businesses that would find more than enough horsepower in
one.

>              The VAX was a great non-RISC instruction set and was widely
> taught between the PDP-11 era and when x86 PCs were so ubiquitous.

And, just think how the VAXen we know an love would perform if all we
did was make the processor using todays technology.  Any guess at how
fast they might be?

> 
> Alpha was outright killed.  Its good stuff has been incorporated into the
> x86 and Itanics, the rest of it is now in a vault which nobody has the
> combination for.
> 
>>>                                     Alpha was killed when it was pretty
>>> much the fastest processor.  Who knows what may have happened if EV8 and
>>> EV9 and beyond saw daylight?  However, Alpha was apparently a real "cash
>>> furnace" which is why it got killed in the first place.
> 
>>Not sure what you mean by a "cash furnace" but if that means burning lots
>>in cost while returning littel in profits I would not only disagree but
>>also have to ask, what of it's supposed replacement?
> 
> I got that term from someone who worked in the Alpha development side of
> things for Digital/Compaq.  He always used that phrase.  It seemed the
> cost of continued development of Alpha, when divided by the relatively
> small number of Alphas actually sold (mostly to Digital/Compaq itself) was
> huge, and an ongoing cash loss for the company.

As I thought, So, what then is the IA64?  A tthis point I would not think
it even possible to ever sell enough to recoup the development costs that
have been poured into it.

> 
> As to its supposed replacement, maybe that was the real reason for 
> switching to a non-DEC/CPQ processor, so some other company would have
> to pay for their own furnace.

Except, that we keep hearing that HP is the one actually pouring the
money into this toilet.  If Intel was actually funding this I would
have expected a shareholders revolt long before now.

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999 at cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>