The need for System/360 emerged from an odd set of circumstances. In the early 1960s computing was finally coming into its own. Some of the wonders that people had confidently predicted right after the war—such as automated weather forecasting—had failed to materialize, but the machines had been put to use every place else, from the monthly mailing of electric bills to the space race. Computers were beginning to revolutionize everyday life. For example, American Airlines was about to unveil its SABRE reservation system, linking ticket agents all over the United States to a master computer in Westchester County, and making it possible for customers to book airline seats without having to put up with the traditional overnight wait for confirmation. We could imagine much broader horizons ahead. One idea that appealed to me was placing computer terminals in America’s classrooms to help raise the standard of education. Meanwhile visionaries began talking about the day when computer power would be delivered to homes just like telephone service and electricity.
Paradoxically, there also was a feeling in the early ’60s that IBM had reached a plateau. We were still expanding, but less quickly than before—in the year Kennedy beat Nixon, for example, we only grew nine percent. As we reached the two-billion-dollars-a-year mark people began to speculate that we’d gotten so big that naturally our growth rate had to fall. But given the bright prospects for computing, that seemed illogical, and I thought it was probably our own fault that we were slowing down. The two separate computer divisions I’d created were competing with each other fiercely, just as I’d hoped, but one unhappy side effect was that our product line became wildly disorganized. By September 1960, we had eight computers in our sales catalog, plus a number of older, vacuum-tube machines. The internal architecture of each of these computers was quite different, so different software and different peripheral equipment, such as printers and disk drives, had to be used with each machine. If a customer’s business grew and he wanted to shift from a small computer to a large one, he had to get all new everything and rewrite all his programs, often at great expense.
The man whose job it was to spur IBM’s growth rate was Vin Learson, who at this point was vice president and group executive in charge of our manufacturing and development divisions. He thought the obvious answer was to simplify the product line, and he asked the technical people to try. But at first this request had very little effect because the engineers of each division were wedded to their own machines. The amazing thing about Vin, apart from the tremendous force of his personality, was his cleverness as a manager. He saw that it was time to break down the rivalry between the two divisions, and he did it by applying a management technique called “abrasive interaction.” This meant forcing people to swap sides: taking the top engineer from the small-computer division and making him boss of the best development team in the large-computer division. A lot of people thought this made about as much sense as electing Khrushchev president, but after interacting abrasively for some months those engineers earned one another’s respect, just as Vin anticipated. Slowly they worked their way around to the idea of building a single computer line that would span the entire market. Vin made this group of engineers the core of a much larger committee called SPREAD—an acronym for systems programming, research, engineering, and development—whose charter was to map out a new product strategy. The SPREAD committee met for a couple of months late in 1961, and when it was slow producing a report, Vin got impatient. Two weeks before Christmas he sent them to a motel in Connecticut with orders not to come back until they’d agreed. That was how the plan for System/360 was born—in the form of an eighty-page report delivered on December 28.
During the year of the SPREAD report I was busy on presidential committees in Washington and followed the technical debates within the computer divisions only from a distance. But by the middle of 1962 enough had gone wrong that I was spending a lot more time at the office. The stock market that May had its biggest drop since 1929, and IBM stock got hammered along with the rest, losing about one third of its value—its first major decline in thirty years. That alone would have been enough to reel me in, and with the 360 on the way, I began to get a hollow feeling in the pit of my stomach about the risks we faced. IBM had always succeeded by making bold moves, but the System/360 plan was dramatic even by our standards.
Vin was the father of the new line of machines. His intention was to make all other computers obsolete—including the thousands of machines on which we were then collecting rent—and to replace them with a completely new family of processors, ranging from little machines renting for $2,500 per month to high-performance giants renting for more than $115,000 per month. The machines would all embody a revolutionary new feature called compatibility, which meant that, despite their great variation in size, they’d be able to use the same software and hook up to the same disk drives, printers, and other peripherals. Once customers shifted to System/360, they’d be able to expand their installations simply by mixing and matching components from our sales catalog. That was good for them, and the benefit for IBM was equally compelling—once a customer entered the circle of 360 users, we knew we could keep him there for a very long time.
From the beginning we faced two risks, either of which alone was enough to keep us awake nights. First there was the task of coordinating the hardware and software design work for the new line. We had engineering teams all over America and Europe working simultaneously on six new processors and dozens of new peripherals—disk drives, tape drives, printers, magnetic and optical character readers, communications equipment, and terminals—but in the end all this hardware would have to plug together. The software was a bigger hurdle still. In order for System/360 to have a consistent personality, hundreds of programmers had to write millions of lines of computer code. Nobody had ever tackled that complex a programming job, and the engineers were under great pressure to get it done.
Our other source of worry was that we were trying for the first time to manufacture our own electronic parts. The electronics industry, after having progressed very fast during the 1950s from the vacuum tube to the transistor, was on the verge of another transformation. The wave of the future was integrated circuits—computer chips that incorporate transistors, resistors, diodes, and so on, all in a single tiny unit. Nobody was using integrated circuits in computers yet, but the System/360 design called for a lot of them. Al Williams argued that even though we’d relied on suppliers to provide the earlier generations of components, we had to manufacture these ourselves. “Whole computers are going to shrink down onto these devices,” he said. “When that happens, do you think we’ll want to be buying them from outsiders? If we’re going to stay in the computer business, we’d better learn how to make these things ourselves.” I agreed, but I’ll never forget how expensive it was to build our first integrated circuit factory. Ordinary plants in those days cost about forty dollars per square foot. In the integrated circuit plant, which had to be kept dust-free and looked more like a surgical ward than a factory floor, the cost was over one hundred fifty dollars. I could hardly believe the bills that were coming through, and I wasn’t the only one who was shocked. The board gave me a terrible time about the costs. “Are you really sure you need all this?” they’d say. “Have you gotten competitive bids? We don’t want these factories to be luxurious.”
Our original intent was to announce the first machines in April 1964 and gradually phase out the old product line by unveiling the rest over eighteen months. Unfortunately we’d miscalculated how much time we had, and the flaws in our existing product line caught up with us a year or two sooner than we’d anticipated. By spring 1963 the old computers were obsolete. We did a technical study showing that while the 360s were going to be better than the latest computers from RCA, Burroughs, Honeywell, Univac, and General Electric, all those competitive machines were superior to our existing line. A number of them offered two to three times the performance of our computers for the same price. Our salesmen were hamstrung—since System/360 had not yet
been announced, and none of them even knew what we were planning, they had nothing to tell customers. By the middle of 1963, sales offices were sending in panicky reports that they could no longer hold the line against the competition. Even though demand for computers increased by well over 15 percent that year, IBM grew only seven percent, our lowest growth since the war.
The only solution was to get System/360 out the door fast, and a number of executives argued that we ought to launch the whole thing at once. This would surely make a tremendous splash in the market. Customers would see how they’d be able to grow with the product line, so we could persuade them to wait for the 360s to be produced instead of jumping to competitors. But there were big disadvantages. For one thing, although the new computers were nearly ready, not all of them had gone through the rigorous testing we required. A bigger danger was that once we started accepting orders, our factory network would be under tremendous strain to deliver every item in an enormously complicated new product line. There wasn’t going to be much room for error.
It was the biggest, riskiest decision I ever made, and I agonized about it for weeks, but deep down I believed there was nothing IBM couldn’t do. And so on April 7, 1964—almost exactly fifty years after my father first came to work at IBM—we staged a product announcement that would have made him proud. To attract maximum publicity, we held press conferences in sixty-three U.S. cities and fourteen foreign countries, while tens of thousands of guests all over the world showed up for customer briefings. A chartered train carried two hundred reporters in New York from Grand Central Terminal to Poughkeepsie, where the main announcement was made. I presented the 360 as “the most important product announcement in company history,” and the visitors were shown into a large display hall where six new computers and forty-four new peripherals stretched before their eyes.
Within IBM there was a great feeling of celebration because a new era was opening up. But when I looked at those new products, I didn’t feel as confident as I’d have liked. Not all of the equipment on display was real; some units were just mockups made of wood. We explained this to our guests, so there was no deception. But it was a dangerous cutting of corners—not the way I think business ought to be done—and an uncomfortable reminder to me of how far we had to go before we could call the program a success.
Williams and I felt that the 360 announcement was the appropriate moment to bring Dick into the IBM mainstream. Until then Vin had been in sole command of the project; now we set Dick up as his peer and divided responsibility between them. Dick was in charge of the engineering and manufacturing side of the business, and Vin was in charge of the sales force. In hindsight I think Vin deeply resented this change, and with reason—the 360 was his product line, and here we were telling him to go out and sell it while we brought my brother in to finish Vin’s job. But to Al and me at the time it seemed natural to divide responsibility between the two men. Although Dick had never launched a major product, he had presided over World Trade’s complicated factory system in Europe. And we needed Vin, with his enormous drive and years of selling experience, out there twisting the tails of our salesmen. Once the assignments were made, Williams and I felt so satisfied with IBM’s course that we announced we would step back from the day-to-day operation of the company. We set up something called the Corporate Operations Board, with Vin and Dick as co-chairmen, and made it their responsibility to run the show.
I honestly thought I’d given Vin the tougher job. The engineering and manufacturing side of the company had tremendous momentum on this project by then, while the sales force was starting from scratch. Not only did they have to turn the tide against our competition, but there was danger that the 360 would alienate a lot of customers. Those used to their current machines were almost sure to balk at the idea of rewriting their software to work with the new line. Vin’s people had to convince customers to make those conversions, while competitors’ salesmen flocked around, saying, “Don’t convert. Come to us. Convert to us.” I was so afraid of losing customers that I called Dick and Vin into my office and lectured them sternly. I told Vin, “If the sales force needs new features or extra software to move those machines, I want you to shout loud and clear and we will produce them for you.” I told Dick, “Be responsive to the sales side of the business.”
But my anxiety was misplaced. We got immense numbers of orders—far more than expected—and even more kept pouring in. Everybody felt euphoric at first, but then the 360 program seemed to get more complicated every month. Competitors found gaps in the new line and started winning away influential customers, so we had to add two new computers to the six we’d already unveiled. We announced an ultra-fast scientific machine to fight off Control Data, and a smaller, special-purpose processor to fight off General Electric. Each new machine required a major diversion of engineering talent. Meanwhile, the effort to write basic software for the 360 line bogged down alarmingly. The more the software was delayed, the more programmers we assigned; by 1966 we had two thousand people working on it, and the cost of developing the software was beginning to exceed the cost of the hardware. We learned the hard way one of the great secrets of computer engineering: throwing people at a software project is not the way to speed it up. A piece of software is a unified thing; if you try to break up the job of writing it among too many people, it takes more time to coordinate them than the division of labor saves. Or as Fred Brooks, the droll engineering genius from North Carolina who led the project, once wrote, “the bearing of a child takes nine months, no matter how many women are assigned.”
There was no way to hide our struggle from customers who were ordering the machines, so I spoke to a conference of IBM users about it: “A few months ago IBM’s software budget for 1966 was going to be forty million dollars. I asked Vin Learson last night before I left what he thought it would be, and he said, ‘Fifty million.’ This afternoon I met Watts Humphrey, who is in charge of programming production, in the hall here and said, ‘Is this figure about right? Can I use it?’ He said, ‘It’s going to be sixty million.’ You can see that if I keep asking questions we won’t pay a dividend this year.” I was only half joking. By the time the 360 software was finally delivered, years late, we’d sunk a half-billion dollars into it alone, making it the single largest cost in the System/360 program, and the single largest expenditure in company history.
Six months after the 360 announcement, the complications of the project were alarming enough that I began holding Monday morning meetings in my office, just as in the days of the Defense Calculator. I disbanded Vin and Dick’s Corporate Operations Board and replaced it with a new Management Review Committee composed of five men: Al, my brother, me, Vin, and Dick Bullen, our organizational expert, who by now was a vice president and group executive. We made shifts like that all the time at the top of IBM, but this one didn’t sit well with my brother or Vin. They knew that I had given them the business to run, and now I was taking back the reins.
Every Monday we’d get together and there would be serious problems on the table. It was not a happy group. Vin and Dick weren’t getting along, and there was also tension between Dick and me. For the first time in our lives he was reporting to me on a day-to-day basis, and those meetings of the Management Review committee were fierce. That committee, which had existed in various incarnations since just before the Williamsburg Conference in 1956, had always been where I evaluated top executives and determined whether they were pulling their weight. Since the success of the corporation was directly related to how this group performed, I judged them harshly and it was easy to find a man wanting, even if his performance in lower jobs had been superior. It was the crucible.
I went after my brother in the same way I’d have gone after anybody else in that position. I remember one episode at the 1964 World’s Fair. Dick and I went out to the Flushing Meadows fairground to look over the IBM pavilion, which had been designed by Eero Saarinen just before his death. The building was dominated by an elevated theater shaped lik
e a giant egg with “IBM” imprinted in bas-relief all over it. Dick and I were touring this exhibit with some associates when the subject of computer circuits came up. I’d been hearing about a new machine, recently unveiled by a competitor, that incorporated an electronic chip, made of silicon, called a monolithic integrated circuit. Such chips are as common as table salt today, but I’d never heard of them, and our competitor was claiming they were far superior to the ceramic-and-metal integrated circuits in System/360 machines. Technology was Dick’s area, so I asked him about it and he was totally mystified. “What are monolithics?” he said. I castigated him for that loudly, telling him he’d better make sure the 360 wasn’t going to be obsolete before it was even delivered. We later found that our scientists had been working on monolithics for years, so there was no real danger. Such incidents drove Dick and me apart.
Father, Son & Co. Page 40
