IBMWatson autobiographythoughts on 777

I seem to recall that even in the 1920s IBM was a very fast growing company. Certainly design improvements to their machines were proceeding apace (in response to Powers machines). Still Social Security increased their market considerably.

Even so the railroads were one of the big consumers of punched cards and EAMs starting at least in the 20s, I believe.

In fact checking IBM's Early Computers (page 4), the first accounting task Hollerith developped his tabulators for was freight accounting for the New York Central Railroad.

I pretty much agree, but ballistics was another major application of machine calculation during WWII. It is worth point out there were scientists, such as Howard Aiken and Wallace J. Eckert, in the 1930s who recognized that they could benefit greatly from automating calculations. Presumably things like monetary constraints prevented them from going further than they did at the time. WWII illustrated the practical value of such work. Also, the short times and high demand for labour during the war probably made machine calculation more attractive then it might otherwise have been spuring demand and developments.

Yes, although they may have been inspired by work at IBM's Watson lab

Well according to IBM's Early Computers (page 21) at the end of 1943 6% of IBM's tabulator inventory (10 000 machines) was non printing tabulators ( so perhaps 600 machines). Also it is mentioned that the non-printing tabulators were far easier to maintain than the later printing models, so there durabiltiy might have led to them surviving on to a much later date. The last new EAM was developped at the end of the 40s but they were used at least until the end of the 60s ( -- Yours Truly, Allan Olley