Talk:IBM 407

From Wikipedia, the free encyclopedia

[edit] Multiplication on the 407

I entered into a discussion with another former tabulating machine technitian concerning the subject of multiplication on a 407. Clearly the old 407 manual does not address this, but it is my recollection that we were able to achieve multiplication through a technique that worked, although very time and resource consuming. It seems to me that the multiplier had to be store and compare against a counter that re-added the multipican for each integer in the multiplier. I don't recall how the cycles were acquired for the operation though.

The discovery of this technique was made by a fellow 685xx techncian in our USAF shop back in 1964. Needless to say, that was 40 years ago and my memory is not very clear after all these years. Does anyone have a better recolleciton about this technique than I do? Any help will be appreciated. YORD-the-unknown 21:13, 12 October 2006 (UTC)

[1] and [2] (unfortunately, what we want is in the referenced documents). From that same old tech 69.106.232.37 01:22, 13 October 2006 (UTC)

Excellent!! Unfortunately, I don't have an active ACM membership so we'll have to wait until someone comes along with one. I'd really love to see the material. :)

Another Columbia reference I found: "Although the 407 is really just a big adding machine, creative use could be made of the control program; for example, as described by Roger L. Boyell in Programmed Multiplication on the IBM 407, Journal of the ACM, Volume 4, Number 4, October, 1957, pp.442-449." [3]

It looks like our top guy learned the technique elsewhere and taught it to the rest of us.

That same guy programmed a 407 as an "alarm clock" one night on the grave shift, when it reached the designated time it started printing full 100 print positions and gang punching 80 column cards, cycle after cycle -- the thing made a hell-uva-lot racket, but I had to wake Cal up because he was out stone cold and the thing wouldn't stop. He hadn't gotten much sleep before starting the shift and didn't want to get caught asleep by the morning crew. Here is another interesting reference [4].

BTW, I do recall that we had one E8 in our bank of 407s but we didn't have a 1620. Ah, the memories! YORD-the-unknown 10:08, 13 October 2006 (UTC)

Concerning the division operation. I'm getting the impression that it was not doable. Here is a link to an IBM page that lists multiplication but not division. [5] It looks like my assumption that if multiplication could be performed that division could also be performed was not a good one. Subtraction had the feel of a bit of a kludge already, so I'm not surprised.