The DEC Jupiter project cancellation [was RE: Stanford's PDP-6 ...

[Rich Alderson]

> From: William Donzelli
> Sent: Friday, June 19, 2009 9:05 AM

>> Tha was not DEC's way. They tended to keep things or offer them to
>> educational establishments. I can't imagine a conversation along the lines
>> of: Q. "Please may we have your PDP-6 to preserve?"
>>     A. " No I want to scrap it"

> In the early 1980s? From every DEChead I have talked to, the early
> 1980s was a time for change in DEC - the company "grew up" and became
> far more corporate. Many say DEC lost their soul in the 1980s.

> And was't Jupiter killed about this time? The PDP-6 could have fallen
> just for spite.

The cancellation of Jupiter[1] was announced at DECUS in May, 1983, the 20th
Anniversary celebration a year and a half later, in December, 1984.  I don't
believe that the cancellation and the PDP-6 disappearance were related.

In point of fact, the DECUS Large Systems SIG had enough influence that they
were able to negotiate a continuation of support for the 36-bit line for a
very long time:  Digital agreed to continue hardware development for 5 years,
and software development for 10.  This gave us the MG20 memory (4MW in two
boxes vs. 3MW in four boxes with MF20); the MCA25 cache (doubled the cache
size); the CI interface, the HSC-50 and RA81 disk drives[2]; and the NIA-20
Ethernet interface[2].  It also gave us TOPS-20 versions 6.0, 6.1, and 7.0,
and Tops-10 version 7.04.[3]

This was a good thing.  It gave the PDP-11, and later VAX, customers an
example of what could be negotiated when those product lines were on the
chopping block.

[1] For those not familiar with the DEC 36-bit line, Jupiter was the code
    name for the follow-on to the KL-10 based systems (2040/2050/2060/2065
    and 1080/1088/1090/1099/1095, defined by operating system and until the
    1095/2065 by some microcode differences).  It was expected to be called
    the 2090, although the sales materials produced immediately prior to
    the cancellation are labeled DECSYSTEM-4050.  It was supposed to be a
    lot faster, but never reached even 50% of the expected speed improvement.
    As much as it pains me to say this, Digital was right to cancel it.[4]

[2] OK, not everything was an improvement.

[3] Possibly 7.03 as well, but I didn't work with Tops-10 until 5 years ago,
    and don't know the history well prior to chairing the DECUS session at
    which TOPS-20 v7 and Tops-10 v7.04 were announced.

[4] Many many 36-bit customers turned away from Digital altogether following
    the Jupiter cancellation, replacing equipment with Sun or HP gear rather
    than VAX (or later Alpha), although a very few were able to hold out for
    the clones from System Concepts, and even fewer for the XKL Toad-1.


Rich Alderson
Vintage Computing Server Engineer
Vulcan, Inc.
505 5th Avenue S, Suite 900
Seattle, WA 98104

mailto:RichA at vulcan.com
(206) 342-2239
(206) 465-2916 cell

http://www.pdpplanet.org/