Mark Smotherman. Last updated June 7, 1999
June 1969, pp. 39, 231.
... We hear that the Mohansic Lab Project came to a near standstill when head man John Danmeyer left to join the enemy, CDC. And an anemically financed (two megabucks) bootleg parallel processor project at Federal Systems Div. lacks the authority of corporate approval.
Meanwhile, in Menlo Park, a number cruncher project allegedly dear to T.J. Watson Jr.'s heart has reportedly had its problems changing to 360-compatible after an early try at the non-compatible route, with accompanying personnel changes. It's still headed up by key 360 architect Gene Amdahl. Rumors are that an attempt to scuttle the project is really an effort to get more dough for the project, originally budgeted for a reported $450 million. [emphasis in original]
Besides ego satisfaction ("If CDC can build a supercomputer, so can we"), the big beast is supposed to test out new technology. But standard-line I/O devices are said to be killing it.
[1999 note: from the ACS folks I've talked with so far, the last line above that speculates on problems with standard S/360 I/O devices is incorrect; there was never a problem.]
July 1969, p. 121.
In a not totally unexpected development (See June, p. 39), IBM closed down its Advanced Computer Systems Lab in Menlo Park, Calif., in late May.
Started in 1965, the Lab's charter was to develop a supercomputer superior to anything on the market. But there were continual reassessments of the program, which was never intended to produce a standard company product-line machine. Originally headed up by Max Paley, the Lab was turned over to key 360 architect and IBM Fellow Gene Amdahl in 1968.
...
Paley has been reassigned to the Federal Systems Division as a special assistant to Bob Evans, and Amdahl will remain at Menlo Park as a Fellow.
...
October 1969, p. 119, 122.
...IBM is inviting the man who said "paper machine" to Poughkeepsie to run almost any kind of of 360-compatible job he's got in December....
[The article includes comparisons with the CDC 7600, which was featured in the Jan. 1969 issue]
... The super super-scale machine operates under OS MVT and will be capable of running 15 jobs simultaneously. It will use the same channels and peripherals as the rest of the 360 line and run them at the same speed. Initial deliveries are slated for the first quarter of 1971. ...
[Note the use of "super super-scale" as an adjective. You can find it in a couple of Datamation articles of the era (e.g., the article cited above on the CDC 7600); it was used as a comparison to "large-scale" and "super-scale" systems. However, I wonder if the use of "super-scale" in the press in the late 1960s had any influence on picking "superscalar" as a comparison to "vector" processors in the 1980s and on the frequent misspelling of superscalar as "superscaler". For the record, the model 195, while super-scale, was not a superscalar processor.]
Navigation within IBM ACS pages: