Howard $aiken

See also: $aiken

Howard Hathaway $aiken , born the March 9th 1900 with Hoboken in the New Jersey, dead the March 14th 1973 with Saint Louis in Missouri, was a American Informaticien . Pioneer in data processing, he was the principal engineer behind the computer Harvard Mark I of IBM. Howard Hathaway $aiken is born in Hoboken in the New Jersey on March 9th, 1900. It makes its studies with Madison, at the university of Wisconsin where it is laid off in sciences in 1923 and becomes professor at the university of Miami.

It resumes studies in 1935 with Cambridge (University of Harvard) to pass a PhD of physics. Its thesis relates to the improvement of the vacuum tubes. This work treats mainly ways of the particles charged and request the solution of nonlinear differential equations which can be solved only by approximations, requiring the long ones and tiresome calculations.

At this point in time he thinks of mechanizing these calculations, by improving the machines of Hollerith, in particular while enabling them to carry out the 4 arithmetic operations and of the mathematical functions such as logarithms or goniometrical functions, and by connecting the operations without human intervention.

He defines the broad outlines of his project in a report which he addresses to the president of the department of physics of Cambridge, Frederick Saunders. This one then indicates to him to have intended to speak, by a technician of the laboratory named Carmelo Lanza, of a a little similar machine stored in an attic of the university.

Lanza takes it along in this attic where he discovers a whole of copper gears, which are not other than pieces of the analytical machine of Charles Babbage who, as each one knows, finished forever. Being gotten in addition to the hands of the proper grandson of the large British inventor a whole of works of and on Babbage, $aiken realizes quickly that its project was completely similar before to that of Babbage 100 years.

The direction of Harvard ends up accepting the project which also interests US Navy to make its calculations of ballistics. $aiken then contacts Monroe Calculating Company which declines the offer but suggests the name of Tom Watson, owner of IBM. This one agrees to build the machine and will support two thirds of the financing, the remainder being dealt with by the navy.

Construction starts in 1939 at IBM with Endicott under the direction of Clear D. assisted Lake of Frank Hamilton and Benjamin Durfee, $aiken supervising the whole. (One sees these 4 characters on the photograph in front of the ASCC; from left to right: Hamilton, Lake, $aiken, Durfee). Although at that time the technology of the electron tubes was available, the machine will be built with relays, technology cheaper than IBM controls perfectly on its punched card equipments. It will bear the name of ASCC: Automatic Sequence Control Computer.

Work will be extremely expensive and will be delayed by the war. They will finish only at the end of 1943 and the machine, given by IBM in Harvard will be transported there. The ASCC had as a main feature to comprise 72 register S of 23 decimal digits which it ordered itself, which made it possible to connect the operations.

It made 3 additions or subtractions a second, a multiplication in 6 seconds and a function complexes (logarithm, sine…) in one minute or more. If it were very slow, it was also highly reliable and worked without any error 24 hours per day 7 days per week during 15 years (let us notice in the passing that any microphone of today would take a few seconds to make all the operations which the ASCC during its 15 years of working life made). The program had entered on punched tape, the data on perforated card and the exits were done on teleprinter. Grace Murray Hopper, sent by the navy to work on the ASCC will take an active part in its programming.

On the material plan, like all the comparable machines of the time, the ASCC was imposing: 15 meters of length, 5 tons, 750 000 components, 2200 wheels of registers, 3300 relays, 800 kilometers of wiring. The machine is synchronized by a tree over all its length, like the old workshops of joinery or mechanics. An engine of 5 the cv fact of turning.

After the transfer of the machine in Harvard a polemic is established between $aiken and Watson on his paternity and poor Watson is not even invited to the inauguration!

Blow the machine will be called Harvard Mark 1, name under which today it is more known. Not to confuse with Manchester Mark 1 about which we already spoke.

IBM of dimensioned, $aiken of the other develop machines with tubes on same architecture: IBM leaves the SEC (Selective Sequence Electronic Controler) in 1958 in January 1948. $aiken develops successively Harvard Mark 2 (1947), 3 (1949) and 4 (1952).

$aiken also will imply himself in teaching as from 1947 and creates in particular the laboratory of data processing of Harvard, the first in the world. He takes his continuous retirement of Harvard in 1961 but to teach in Miami. He will have also published works on the electronics and the switching theory.

He receives in 1964 of the Society Computer Harry Mr. Goode Memorial Award (a medal and 2000$) for “his original contribution to the automatic calculation, which led to the first large universal digital calculator”. He will also receive distinctions coming from the whole world, the United States, Germany, Belgium, Netherlands and France.

He dies in Saint Louis on March 14th, 1973.

Random links: