> By 1979 360/75s were 10+ years old.

By 1979, 360/75s were 13-14 years old. TUCC in RTP, NC got one of 
the very first early in 1966. I was told in 1967 by one of the CEs
in residence at TUCC that it wasn't the first one out of the barn, 
but it was nearly so. We knew that NASA had ordered five, and you
can find references on the web that claim NASA installed them in
the Fall of 1965, but which ones in the manufacturing sequence the 
actual NASA machines were was a "we can't talk about that" issue 
at the time. For one thing, I think NASA got them early, before 
first-day orders from real customers were shipped. Another issue
is that their delivery was rushed (already being later than IBM
had promised) in order to prevent NASA from giving up on IBM 360s
and using CDC 6600 machines exclusively. Additionally, the 360/75s
sent to NASA had an RPQ modification to provide a real-time GMT
clock facility ("Store GMT" was the instruction name, I think).  

The five Model 75 machines that NASA used for the Apollo project
were no longer needed by 1973, and were turned over for use as
development machines for the Space Shuttle software development.
They were used for that job until January 1982. JPL in California
picked up two Model 75s from NASA (one of them had the "Store GMT"
RPQ, which turned out not to be sufficient for JPL's purposes, and
they had to get IBM to extend it with more bits), and then JPL did
something surprising even then: they BOUGHT a third 75, which they 
used until August 1983. 

That August 1983 date is the latest that I have ever heard that a
Model 75 was actually in use. Univ. of Waterloo got rid of theirs
(as junk, but valuable junk because of the gold in the SLT cards)
in late 1979 or early 1980 (it was turned off for some number of
months before they actually physically got rid of it).

According to this web page:
http://www-03.ibm.com/ibm/history/exhibits/mainframe/mainframe_PP2075.html
the Model 75 was "withdrawn March 15, 1977." It also indicates 
"... the Model 75 had a monthly rental range of $50,000 to $80,000, 
and a purchase price range of $2.2 million to $3.5 million. 
Deliveries began during the fourth quarter of 1965."

That purchase price range seems a little high, so it is possible 
that what they are referring to is the 1977 purchase price for a
2075, instead of the 1965 to 1968-era price (before unbundling).
But, surprisingly, that monthly rental price range (presumably
referring to the low-end 2075-H [256 KB] thru the high-end 2075-J 
[1,024 KB]) seems a little low. I have original documentation for
the exact equipment prices [as published by IBM] as of Aug. 1968,
and used those figures to make my calculations. In addition, I 
have original documentation for what TUCC was paying IBM (after
their 40% educational discount); each set is consistent. Thus, I
can only conclude that the IBM web site price figures represent
lower monthly rentals at end of equipment life, but also a higher
outright purchase price 12 years after its original announcement.

--
WB    

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to