Forecasting or Reliability, Serviceability (RAS) and Rental Cost
Having worked most of my professional life at IBM connected with Manufacturing and Machine Development, it failed to me to realize how much the existence of any equipment depended on the forecasting of that equipment. It does not matter how brilliant, how clever or whatever reason around these justifications existed, if it does not meet the forecasting expectancy of who foot the bill: IBM. I could say that any successful or responsible company which is paying by its own means such expense justification does the same. Forecasting when deciding on bring to existence a product at IBM is called Market Planning. The Market Planning Manager for the 1401 was Sheldon Jacobs, and from his speech at the Computer History Museum, he stated that Market Planning or simply Planning, has three main aspects:
What users look like?
There is a basic question that very seldom, if any, is questioned by historians and anlysts but which it is basic to IBM decision on creating a machine: it has to be a consumable product.
They used a lot of punched cards, Keypunches, Punched card input/output and Unit Record equipment. IBM didn`t sell its equipment, it rented them and the punched card had to come from IBM. From 1930 to 1950, some 25% of profits came from punched cards. The US Supreme Court ruled this out in 1936. Later on, in another case, heard in 1955, IBM signed a consent decree requiring, amongst other things, that IBM would by 1962 have no more than one-half of the punched card manufacturing capacity in the United States. Tom Watson Jr.’s decision to sign this decree, where IBM saw the punched card provisions as the most significant point, completed the transfer of power to him from Thomas Watson, Sr. In 1955 a box with 1000 punched cards cost US$.99 or roughly a dollar and that was the basic reference to compare business interest, i.e., a dollar was a point and the rental of the machines were all defined in terms of points. Shel Jacobs was very skillful quantifying that and after extensive research, he came to a conclusion that the prospect customer of a “small” computing system capable of replacing what such customers computational needs looked like, would run in the 2000 points break even point. So he set out 2500 points as reference to the 1401, or US$2500 dollars monthly rental. He underestimated initially that it would sell 2000 machines in its life cycle, raised it to 5000 after the enormous response from customers and the 1401 ended up selling 15000 machines in total. This enabled IBM to have the necessary money to face the IBM 360 challenge, which is considered fairly the same cost of the development of the A bomb in 1945 and I would imagine that it is comparable with the costs to have the man on the moon in 1969. William Rodgers in his best seller Think, discussed in more detail bellow, stated that the IBM 360 development costs in 1967 dollars $5 345 000 000, Such was the money the 1401 brought to IBM.
How did the punched card fit into the operation? Lets take a look on the typical sequence the punched card goes through:
1 – Key punch
3 – Sort
4 – Collate
5 – Calculate
6 – Tabulate
7 – Summary punched
Shel Jacobs found out which were the most complex operations and what they were all about and went to these customers to ask them what they wanted. “More of the same” they replied. The control panels were carefully studied and Frank Underwood came to conclude that with some 3000 memory positions he could emulate any such control panels, provided that one of his most successful insights were put in place: The variable length instruction. Machine programs previously tended to use only a portion of the memory position, because they were fixed length, leaving portions of the memory empty, wasting precious and costly memory positions. The idea of resetting each instruction with a “word mark”, or a character, enabled to make full use of the empty space, what meant in practical terms more memory for more sophisticated programs. After Frank Underwood findings and advances in programming, he came to the conclusion that a 4K positions memory would be enough, and so it was defined that would be the size of the memory of the 1401. Frank also introduced “user-friendly” acronyms for the instructions, such as “A” for adding, “M” for multiplying “B” for branching “E” for editing and so on. It ended up with two languages, autocoder, when using tapes and symbolic language when using only punched cards. RPG or Report Program Generator was also an innovative feature and together with the Input.Output Control System (IOCS) saved time and trouble of programming. (Good Bye Control Panels!…) The control panels were hell and prone to errors to be programmed and they not only were eliminated, but replaced with a much more accessible format to make the machine to do what was wanted or needed.On top of that, everything was to be transistorized! Thomas Watson Jr foresaw where the electronics technology was heading to and simply discarded tubes and ordered that everything from them on should be transistorized! For those more familiarized with main frames, take a look on the 1401 Reference Manual
One thing that also never occurred to me is that behind all that charm connected to computers which is always thought and linked with most complicated things as it was the case of the initial entry of IBM into the world of computing, namely Mark I and the IBM_SSEC,The real question was what really mankind needed as a whole at an affordable cost. Anyway, those initial sophisticated machines became instrumental to the development of the first family of “commercial” albeit really “scientific” computers, the IBM 700 family.How did this happened?
Thomas Watson Senior & Jr
IBM, is the brainchild of Thomas Watson, although a creation of Charles R.Flint, who invented the system of combining corporations into trusts. Some consider Thomas Watson as the Henry Ford of computers, but IBM was really a creation of his son, Thomas Watson Jr. His father grew up and lived with a line o products, namely, meat slicers grocery store scales, time clocks and primitive tabulators, all promoted as “business machines” as it can be seen bellow:
In 1911, Charles F. Flint, a noted trust organizer, engineered the merger of Hollerith’s Tabulating Machine Company with two others – Computing Scale Company of America and International Time Recording Company. The combined Computing- Tabulating- Recording Company (C-T-R) manufactured and sold machinery ranging from commercial scales and industrial time recorders to meat and cheese slicers, along with tabulators and punched cards. Based in New York City, the company had 1,300 employees and offices and plants in Endicott and Binghamton, New York; Dayton, Ohio; Detroit, Michigan; Washington, D.C.; and Toronto, Ontario.
As you can see in the Amazon reviews for the William Rodgers book, and I quote:
“Williams Rodgers was a gifted writer for many newspapers and magazines, including the New York Herald Tribune and Harper’s. This book was a huge best-seller when it was published in the late 1960s and very accurately describes the early dominance that IBM had in the computer industry. Although at the time IBM resented the close look its company was given in these pages, later on a member of the Watson family said it was the best biography of the company’s founding genius, Thomas Watson. Try to track down an out-of-print company, as it makes for fascinating reading, still today.”
I entered to IBM Brasil in December 1970 and my first job was to do the layout of the Sumare Plant and to install it, part from IBM Rio de Janeiro, part from the provisory installation in Campinas, SP. An old-timer IBM’r was sent to help us and he had one copy of this book. When he went away he left the book among other things on his desk and I took it and have it to this day (2016). I read the book avidly and the saga of IBM described in the book fascinated me and fueled and feed my youth dreams to go to the USA and see how it really was. I moved from Industrial Installation department to the Product Engineering department to pursue my dream, what was permitted as some sort of prize or recognition for the good job I did in solving the layout problem of the Sumare Plant.
Incidentally, the very first day I came to work at IBM I was taken to visit the Sumare Plant, which was empty, except from a 1401, which was deposited there, waiting to be installed.
The account William Rodgers does about how Tomas Watson Sr got involved with scientific calculations is one of those things. It explains how IBM got involved with Columbia and Harvard, namely, how the Think motto was instrumental in involving IBM with heavy players in the scientific arena. It was, and it is, obvious, that to Think constructively, to become expert in one’s work, one has to be educated. And this was strongly encouraged within IBM anywhere it did business, as a consequence of Thomas Watson Sr. Think perception. Although a little bit poisonous the observation of William Rodgers that, and I quote, “Watson felt little need for academic learning that might have broadened his judgment, or allowed him to view the world with something more encompassing than a copy book maxims and bromidic verities.”, he was enormously helped in to that by Benjamim D.Wood, and I quote once again from this book:
“Watson had the good fortune to have the vision of his own corporation substantially widened by a diffident, scholarly intellectual at Columbia University. It was good fortune for which he was ready, and to which he was immediately responsive. His benefactor in what amount to a step in Watson’s own education was Benjamin D.Wood, an US Army psychologist in World War I and, in 1928, a thirty-four year old professor of Collegiate Educational Research at Columbia.”
Ben Wood’s doctorate dissertation on his doctorate thesis revolved around E.LThorndyke, which simply put states that “whatever exists at all, exists in some amount”, and applied it to the measure of human intelligence.
How IBM got involved with scientific calculations
If you read the two articles above involving Ben Wood and Columbia, you will like to know the human history behind the scenes, as told by William Rodgers, and I transcribe it from the book above:
“He was appointed assistant to Dean Herbert Edwin Hawkes of Columbia College, with the task of advising third-year students entering business, journalism, and law. To do this, he needed some form of prediction of their probable performance in professional schools. Because of previous experience with Thorndike, who was called “the big chief” at Columbia-and who devised the measurement tests which still bear his name-and because he was appalled by the fact that only two out of every seven boys entering college ever graduated, Wood began to devise testing and scoring methods that turned up some truly astonishing findings. In particular, Wood demonstrated a disparity in the quality of education so extreme that seniors achieving the highest possible test scores in some colleges were still below students who got the lowest scores at, for example, Haverford College. His findings seemed incredible, and Wood was pretty roundly denounced as a fake for “mechanizing” measurement best left to human evaluation.
Dean Hawkes weathered the criticism, however, and supported Wood in his investigations. Grants from the Carnegie Foundation, the Commonwealth Fund, and a half-million dollars from the General Education Board subsidized his efforts, and before long the New York Board of Regents and the educational system in Pennsylvania began to reevaluate their objectives in relation to Wood’s revelations.
The monumental task of recording and processing test results became a physical impossibility. Handling 35,000 tests at one time costing five dollars each to process was too costly even with large and generous grants. When Wood had millions of tests to deal with, he was harassed beyond anything even his own tests could measure. He was given floor space in Hamilton Hail, which he furnished with wooden boards and sawhorse tables scrounged from secondhand stores and lumber yards, and staffed with “two acres of girls” reading and classifying test answers. A labor problem arose when the girls numbed by the tedium of the work, gave vent to their wrath over snagging their stockings and clothes on the improvised furniture. Burlap was tacked to splintered surfaces and the uprising quelled. In desperation intensified by the knowledge that the work was enormously important, Wood culled from directories the names of chief executives of ten corporations in the equipment manufacturing business and sent out a call for design and engineering help. Nine answers were “brush-offs from the secretaries of third vice-presidents.” The tenth was a crisp telephone call:
‘I’m Thomas Watson. I’m very busy and can spare only an hour. Be at the Century Club promptly at twelve; I have an engagement at one.’Wood had scarcely said a word, partly out of shyness but largely out of inopportunity.”
The meeting lasted almost 5 hours and started one of the most fruitful relationships ever between the industry and academia.
Watson was eager to get government contracts, specially military and Ben Wood, based on his thesis around Thorndyke ideas, told what was music to Watson’s ears: According to Thorndyke and his philosophy of quantitative measure, Watson should bear in mind that in psychology and elsewhere, whatever existed at all exists in some quantity. If it is zero, it is nonexistent. But all the existing aspects of earth, science, nature, life, intelligence could me reduced to measure. Money and inventories, the hours a man worked, and the rice and profit of goods where not the only things that could be manipulated quantitatively. Everything from virus to super-galaxy – from microcosm to macrocosm – was a matter of quantity. The only way man could ever learn all the things needed to know to keep civilization ahead in the race between education and catastrophe was to recognize que quantitative basis of all phenomena. The machines of IBM could extend all measure whether it was for the doctor, who took a temperature, counted a pulse, took a blood sample, counted cells, or for any other field of endeavor in which the need for knowledge preceded judgment or activity. In summary, there is no aspect of life to which these IBM machines cannot make a basic and absolutely essential contribution. According to William Rodgers, the connection was made this way.
Ben Wood was the first PhD to become IBM’s consultant for life and access an enormous amount of equipment and money to do whatever he deemed right to do and this is how Columbia became one of the most advanced computer users for scientific research. It is very interesting to note that from 1927 to 1932 Ben Wood secured a couple of million dollars in subsidies from the typewriter industry, in which IBM itself later became dominantly involved. His interest was to demonstrate the value of the typewriter as a learning machine. Wood always made an explicit distinction between a teaching machine, used to convey facts and ideas an a learning machine, “an instrument to stimulate creativity and independent, self-initiated and success motivated learning, and to expedite absorption and comprehension..” He didn’t, as far as I know, said, but he could have said that this is basically the difference between training and education as I would discover later in life.
The results can be better perceived at the Columbia site entry on the SSEC
and the Columbia
Thomas J. Watson Scientific Computing Laboratory, 612 West 116th Street (1945-1953).
Two perfect examples of a scientific calculation and what such users looks like, when Government sponsored
I – The Battle of the Atlantic
“Le bon Dieu est dans le détail” (the good God is in the detail) is generally attributed to Gustave Flaubert (1821–1880) and although later on it was reversed to “The devil is in the detail”, if there is somewhere where this happens is when at war. And such was the case of the Battle of the Atlantic. The big picture can be seen above, but William Rodgers in his aforementioned Think publication gives an account of a details that, depending of which side you look at it, it can be either God or devil’s detail. Let’s take a look and figure out how this brings up comprehension on what it was about scientific calculation from the point of view of perhaps its more important user: The US government
As it was told by William Rodgers, in Think, 1969
“Long before Pearl Harbor and the United States declaration of war, merchant tonnage losses of U.S., British, and Canadian bottoms increased to catastrophic proportions. Tankers, in particular, were torpedoed and left in flames within sight of American coastal citizens. The ports at Halifax and the Gulf Coast were graveyards for vessels. With war declared, the situation grew so ominous (although at the time the real danger was not made public) that German submarines along the Great Circle route of the North Atlantic operated nearly at will.
Working in deepest secrecy, Dr. Wallace J. Eckert (not to be confused with Dr. J. Prosper Eckert co inventor of ENIAC) and his staff at the Naval Observatory were charged with carrying out an operation that meant, almost literally, the difference between life and death to the Allied cause, since on it depended the use of the sea lanes to Murmansk and the United Kingdom. Before such technological miracles as radar and sonar were extensively in use, priority was given to means more readily at hand to disperse the murderous submarines which had sent to their deaths half of the pre-war U.S. maritime tanker personnel and the ships they manned.
Effective air assault against submarines was inevitably delayed by navigation techniques requiring as much as thirty minutes to determine a ship’s or a submarine’s position, a preliminary requisite to a radio call for help. Manual calculations, even with available tables and a sextant, were too time-consuming and cumbersome, leaving supply laden ships exposed to torpedoes and destruction in the interim.
Working from Professor Brown’s tables, their accuracy confirmed by the IBM equipment at Columbia, Dr. Eckert, a supervisor at the Observatory named Jack Belzer, and a group of young especially trained women began calculating and producing nautical almanacs for air and sea navigators. Limiting the operation to cover a ten-degree band of the heavens over the North Atlantic-it would have required a million pages of condensed type to provide complete almanacs for the navigable waters over the full north-south range-they produced the printed calculations that constituted the first and oldest scientific computer output in the world. It had been done, experimentally, at the old Pupin computing laboratory at Columbia, where fire control equipment calculations for the B-29 aircraft were later done under subcontracts to the General Electric Corporation.
The almanacs were modified to carry very small slugs of type, printed a line at a time. The type was so tightly condensed that every other digit was printed in the initial operation, then the platen on the machine was shifted one-half a notch to open up alternate spaces for the remaining half of the data on each page. This was done to compress the data and reduce the bulk of the document. Calculations were related to specific dates, and production schedules of the almanacs sometimes left only a few hours during which their delivery was absolutely crucial to navigators in the spotter aircraft, in the assault planes, and on the ships. In Washington, planes stood by to fly them to waiting navigators. Eckert and the staff were never more than a week ahead of delivery. But with the data, navigators on North Atlantic patrol and on ships in transit could determine a fix often in a single minute after sighting a submarine, and radio its position to every craft within range. Corvettes from Canada, destroyers, and anything capable of carrying a gun or a depth bomb could then converge on the spot at maximum speed.
In a matter of weeks, loss of lives and tonnage in the North Atlantic diminished; in time, and with new sensing technology, the sea lanes were brought under Allied control.
It was the sequence of events started by the cooperative Watson, in response to the requests of Ben Wood and Wallace Eckert for IBM machines to pursue their work and scholarship, that led directly to these accomplishments. Watson’s Astronomical Computing Bureau at Columbia, established after Eckert had read the articles written by scientists at England’s Nautical Almanac Office at Greenwich, was the seminal instrument for giving birth to the new machines to measure the phenomena of the universe.
The newest and most apocalyptic phenomena to which Watson’s machines were applied were the nuclear fission and atomic bomb projects; Eckert and Grosch labored on these, not altogether certain of the objectives of their research until word filtered back to them about the Manhattan Project and the awesome accomplishments of Dr. Robert Oppenheimer and his associates, who were dispatching instructions by mail postmarked Los Alamos, New Mexico.
Eckert and Wood remained in close touch with Watson until age, failing health, and his son’s ascension elevated him to emeritus status. Following his years at the Naval Observatory, Eckert was preeminent among scientists of the world in his field, and thus precisely suitable as Watson’s choice to direct, at $30,000 a year, the Thomas J. Watson Laboratory, endowed and established at Columbia after the world entered the nuclear age. A good and true scientist, he served in the post for twenty-two years and found his own heaven where mystics always said it was, among the galaxies and the stars.”
If you go to the pointers above you will have a very goo overview of not only computer genesis, but the IBM computer genesis history. This is the missing link to got to the quantum leap that the System 360 was not only to IBM, but to the world. If you want or need to know what the Von Neumann concept is all about, you should take a careful look at the 1401 Reference Manual because it is all there. Frank Underwood made a perfect use of the Stored Program invented by von Neuman. It is curious his remarks about von Neuman when he was in Endicott acting as consultant Von Neumann came as a consultant to answer questions, deep mathematical problems, he would sit back think he was sleeping, spring up and this is the way it is.
By 1949, the General Picture for such kind of scientific computing looked like that.
III) Announcement Preparation