ON JUNE 6, 1954, DUDLEY BUCK AND JACKIE WRAY WALKED DOWN the aisle at Unity Church in her hometown of North Easton, Massachusetts. They had set the date one week after Jackie named the cryotron, over a dinner with Otis Maxfield, the minister at the Wilmington Methodist Church, whom they wanted to conduct the service.
Schorry Schick, Buck’s lodger and future brother-in-law, was his best man. They had grown close in their months together, in spite of tensions from the neighbors—some of whom were still a bit paranoid about a German living in their midst in the years following World War II.
Jackie’s sister Gwen was maid of honor and Nancy Whitman, the mutual friend who had introduced the young lovers, was bridesmaid. George Frideric Handel’s “Largo” played throughout the service in the pretty little church.
Glenn Campbell was a notable absentee. He had gone back to Washington, DC, shortly before the service to stay with his brother, fearful that he would get in the way of the newlyweds. No matter how much they protested, Glenn refused to listen. He grabbed his bags, and the $150 Tosi accordion that Buck had just bought him, and caught a bus home.
It wasn’t just that he wanted to get out of their hair. Letters from his brother Bill were filled with tales of parties and dancing, which sounded a lot more fun than anything Wilmington had to offer. Yet Glenn never lost touch with his foster father, and to his dying day kept the last letter he ever received from Buck in the drawer of his bedside table.
Buck had been making preparations for married life. The Plymouth sports car was on its last legs: every other week it conked out on him. He would fix it himself using parts he found in Goldie’s Junk Yard in a town nearby. He rigged a chain to one of the sturdy pine trees by the driveway that he would use to lift the engine block off its mount and lower it back again.
As his wedding day approached, Buck splashed out $645 on a maroon 1950 Studebaker Champion, a bullet-nosed two-door sports car. Straight after the wedding reception, Mr. and Mrs. Buck jumped into the car and set off on the long drive to Santa Barbara so the young scientist could introduce his wife to his father, his Grandma Delia, his brother Frank, and the rest of the Buck family.
Although it was their honeymoon, Buck was never one to let an opportunity go to waste. They stopped off in Washington on their way across the country so Buck could pick up some instructions from one of his NSA handlers to carry with him to Magnavox, one of the agency’s West Coast contractors. Before the newlyweds hit the road again, the NSA gave him his wedding present: official orders to go back to the National Bureau of Standards lab in Corona, California, between the dates of June 9 and 21. Buck does not seem to have visited the Bureau of Standards computer lab at all on this trip, but he certainly claimed back his gasoline money for the whole round trip to California as an expense. Thus, American taxpayers picked up the bill for the Bucks’ honeymoon.
The Buck newlyweds drove through Virginia along Skyline Drive that runs through the Blue Ridge mountains, then through Tennessee, Georgia, Alabama, and on to New Orleans and then Houston before trekking west to Santa Barbara. The journey took about a week.
Jackie felt instantly at home in Grandma Delia’s expansive bungalow. She was astonished how big it was inside relative to its humble exterior. Compared to her own “Waspy” New England family, Jackie found the warmth of the aunts, uncles, and cousins a breath of fresh air. There was an ease to their laughter she had never encountered before. And the entire family was impressed by her in turn.
After a few days of laughing, joking, and eating, the newlyweds turned around and headed for home again. By the time they got back to Wilmington, the house was empty. Not only had Glenn gone, but now Schorry had left too. Buck’s sister Virginia had been sent back from Berlin to a new posting in Washington; Schorry had packed up as soon as he got the news, and rushed south to meet her plane.
They gave the place a lick of paint and stripped down and repainted an old China cabinet that was being thrown out by one of the neighbors. Buck built a bookcase along one of the living room walls. The sofa and armchair were reupholstered, and proper curtains were fitted to every window for the first time. It was a happy, cheerful place where they started to get to know one another properly. The two young lovers had only been together for sixteen months before they got married.
Once they had the inside of the house looking up to scratch, Buck set to work on the garden, begging some rich, loamy soil from a local dairy farmer, and then planting a lawn, followed by some yew trees, laurel bushes, and rhododendrons that he neatly placed around the edge of the house.
He was preparing their home for the family they both wanted. In the lab, meanwhile, Buck’s other baby was taking shape.
AT THE HEART of the sprawling campus of MIT sits Building 10—the pillar-fronted palatial construct topped with its famous “great dome” modeled on the Pantheon in Rome.
On the third floor of this imposing edifice, Buck—and a handful of other young electrical engineers and physicists who were working on different computer technologies—were given some lab space.
Wires and cables flew everywhere around the room, firing up strange whirring machines on the workbenches. Buck’s experiments, conducted at temperatures close to absolute zero, stood out from the crowd.
Liquid helium would arrive from the physics department in giant dewar flasks—the laboratory equivalent of thermoses. Buck was trying to find the right combination of metals to make his cryotron work. He was taking delivery of wires made from chemicals that most of the staff had only seen before as symbols on the periodic table of elements.
Initially he had tried to make the device using the cheapest super-conductor: lead. He struggled to find lead that was pure enough to perform the task, however. The thickness of the wire was also a consideration. He ordered tantalum wire of 0.007" thickness, and 0.003" niobium wire in one-thousand-foot rolls. At room temperature, wires that are thinner and longer have higher resistance. It turned out that this was not the case at cryogenic temperatures: a superconductor was a superconductor.
After learning of his experiments, General Electric agreed to donate a small sample of an even rarer superconductor called rhenium. Only a few years earlier, rhenium had sold for about ten thousand dollars a gram, but it had become more freely available after World War II.
Buck’s work was creating a stir, and the word cryotron had entered the lexicon of Building 10. Within weeks of Jackie coining the phrase, there were master’s-degree students choosing to write their theses on the tiny gadget.
Buck had started to explain to Howard Campaigne, Joe Eachus, and Solomon Kullback at the NSA how his cryotron could hold the key to faster, smaller computers. If it could be perfected, the new device could be applied to almost all of the problems they were trying to solve in terms of codebreaking, large-scale data processing, and possibly more aggressive military applications.
Still, it was not much to look at. The cryotron was one straight wire, about one inch long, with a second wire wrapped around it. Both wires were made from superconductors but using different metals that hit their magical state of superconductivity at different temperatures, which added a little to the complexity of the early experiments.
By passing a current through the second wire Buck created a magnetic field that would see the straight wire running through the middle switch from being a superconductor to a resistor. Once it did that, he could create a computer circuit: the cryotron could flip from on to off, yes or no, one or zero.
There were a lot of variables to contend with before he could get there. He had to test the reliability of the two metals that would comprise the device. He had to make sure that the temperature in the flask did not get affected by heat from some of the other apparatus needed to make it all work. There was no textbook to work from or improve upon; it was a new field of experimentation. Trial and error was the only way to get to the right answers.
A miniature production line started. The lab secretaries, who had grown accustomed to being asked to thread little magnets onto wires to make parts for the Whirlwind machine, now found themselves wrapping tiny pieces of wire around other tiny pieces of wire. After a few weeks, Buck built a little spindle machine to wind the cryotrons for him in an attempt to bring more consistency to his experiments and to spare the secretaries’ grumbles.
By November 1954 the twenty-first incarnation of the cryotron appeared to work properly and consistently as a switch. One wire was made of tantalum, the other niobium.
The very next week Buck went to Washington to tell Eachus and the rest of his NSA handlers all about his new device. Things started to move quite rapidly. The cryotron was a long way from perfect, but Buck was confident that he understood its shortcomings and had credible thoughts on how to tackle them.
Academically, Buck’s career was also starting to progress. In January 1955, two months after he got the first cryotron prototype to work, he finally got his first teaching job, giving classes on computer control components at Northeastern University’s evening school. He still worked at MIT by day, and lectured at Northeastern on a handful of nights a week.
He then got another job offer, this time from Arthur D. Little, the chemical analytics company (which has since evolved into a management consultancy). Buck was offered a retainer of two hundred dollars a month. His task? To make more cryotrons. Little, which was closely tied to numerous government research projects, wanted to pursue Buck’s idea commercially. They suggested he work for them on a part-time basis, to make sure they were on the right track.
It was an easy offer to accept. The company had one thing in abundance that Buck sorely needed: liquid helium. Little had built the first commercially available helium liquefiers, but at a princely sum of twenty-three thousand dollars such a device was somewhat beyond his research budget.
The helium liquefier in the MIT physics lab could produce twenty-seven liters of liquid helium an hour. Arthur D. Little’s machines could churn out only four liters an hour, but they had dozens of them. Working with Little could see Buck get his hands on sufficient quantities of liquid helium to experiment on multiple ideas simultaneously.
Helium gas supplies had started to pose problems for Buck by this point. S. C. Collins at MIT had been receiving stocks of helium for free, directly from the US Department of Defense, for his initial experiments, but Buck’s substantial requirements had put a stop to this pro bono support of science. Collins had been instructed that any future supplies of helium gas be ordered with the Boston Naval Shipyard at a price of $4.20 a cylinder.
Buck’s new arrangement with Arthur D. Little was fortuitously timed. Although he had yet to announce the cryotron publicly to the world, the rival scientists operating as part of the NSA’s industrial cooperation program were well appraised of his progress. He was happy to explain the concept of the cryotron and the broad outline of his work to almost anyone who enquired about it, whether they were commercial rivals or academics at other institutions.
Although a flood of new commercial consulting offers appeared, Buck started turning down all new approaches to work on the cryotron. In one rejection letter to a prospective employer, Buck explained the progress he was making with the device he had built in the Arthur D. Little lab. “The work is going along smoothly,” he wrote. “We have three small flip-flop clock circuits which operate at about 1 kilocycle per second and are trying to work with finer wires and a variety of materials and temperatures in order to push the speed up toward the megacycle region. We are being forced to learn a lot about metallurgy in dealing with fine wires. The prospects for a major breakthrough in this computer field with this new component look as good as when I visited you.”
With a queue of America’s new corporate titans desperate to hire Buck, MIT eventually gave him a teaching job in July 1955, appointing him as an instructor in electrical engineering on a salary of $3,420 a year, which worked out at about an extra seventy dollars per month.
A fortnight later he turned down a job offer from the industrial conglomerate Westinghouse, which offered to pay him a colossal starting salary of $8,700 a year. It was an “attractive offer” for a job “doing almost exactly what I hope eventually to do,” Buck wrote back. “But I have weighed the relative advantages of the two possibilities and I feel it is to my advantage to invest another year in academic work at MIT. I am deeply grateful to you for your time and interest in interviewing me.”
Buck had committed to MIT, and MIT had committed to him. Now that he was finally a proper member of the department, he published a paper on the cryotron to circulate among the teaching staff. The abstract of the paper spelled out clearly what he had just created:
The study of nonlinearities in nature suitable for computer use has led to the cryotron, a device based on the destruction of superconductivity by a magnetic field. The cryotron, in its simplest form, consists of a straight piece of wire about one inch long with a single-layer control winding wound over it. Current in the control winding creates a magnetic field which causes the central wire to change from its superconducting state to its normal state. The device has current gain, that is, a small current can control a larger current and it has power gain so that cryotrons can be interconnected in logical networks as active elements. The device is also small, light, easily fabricated, and dissipates very little power.
Superconductors were considered highly experimental. There was a great deal of skepticism about the claim that superconduction equated to perpetual motion. Before even spelling out the nature of his own invention, Buck felt compelled to lay out a defense of the broader field of cryogenic research: “Below the superconductive transition the resistivity is exactly zero. That it is truly zero is vividly demonstrated by an experiment now in progress by Professor S. C. Collins at M.I.T. wherein a lead ring has been carrying an induced current of several hundred amperes since March 16, 1954 without any observable change in the magnitude of the current.”
The paper was dated August 22, 1955, so Collins’s experiment had been running uninterrupted for seventeen months. Resolving that he had laid that doubt to rest, Buck proceeded to explain how he had been able to wire large groups of cryotrons together.
The device worked perfectly as a computer memory, Buck explained, and could also be used to control other types of computer memory and work with filing systems to code and decode where data had been stored. It could add, subtract, multiply, and divide. It could be used in electronic logic circuits; it could operate as an amplifier or with a power control circuit; it could be used to convert analog data received from real world inputs into a digital signal, and then convert it back again.
In different groupings and configurations, these two little pieces of wire suspended in helium could be used to create every type of circuit needed to build a computer. While most computers still occupied whole buildings, this one would fit in a box that was just one foot square.
The cryotron in its present state of development is a new circuit component having power gain and current gain so that it can be used as an active element in logical circuits. It is easily and inexpensively fabricated from commercially available materials and its size is small. Extrapolating the volume occupied by the present experimental circuits to larger numbers of components indicates that a large-scale digital computer can be made to occupy one cubic foot, exclusive of refrigeration and terminal equipment. The power required by such a machine extrapolates to about one-half watt, once again excluding refrigeration and terminal equipment. The reliability of cryotron circuitry is not known, but it is anticipated that operation in an inert helium atmosphere at a temperature near to absolute zero where chemical activity and diffusion processes are essentially stopped promises a high degree of reliability. The circuit noise level is similarly not known, but due to the low temperature, very little thermal fluctuation noise is anticipated. The device is at present somewhat faster than electromechanical relays, but far slower than vacuum tubes and transistors. A program is under way to increase the speed.
Dudley Buck had invented a whole new field of physics and electrical engineering. Though his name disappeared from public view after his death, his legacy runs deep through the veins of the NSA, and computer science, to this day.
As David Brock from the Computer History Museum in Mountain View, California, explains,
Is there someplace else where some person may have made a superconducting switch? Probably. But Dudley Buck is the person who got this whole field of superconducting electronics going.
Since Dudley Buck’s time, it has been a constant dream of the NSA to build its frozen supercomputer. The NSA has never given up on the promise. If you can make a computer that uses almost no power, you can have these gigantic systems. The NSA needs all this capacious memory. They need the sheer processing power. You can see the problems they get into with things like this data center they built in Utah—the power in that thing is unbelievable. They have had fires in there. Power limitations are very real.
The promise of superconducting electronics, where it almost uses close to no energy per unit of computation, means they don’t need to care.
They have tried consistently. What they have failed to do is to invest, and to persuade industry to invest. But they try, to this day. There’s a multi, tens-of-millions of dollars project now being led by IARPA [Intelligence Advanced Research Projects Activity], the intelligence community’s research and development agency, to build a superconducting computer. It’s built of cryotrons. They call them something else—tunneling Josephson junctions. But those are cryotrons.
BUCK HAD BEEN married for a little over a year by the time the cryotron was starting to create a fuss. In the summer of 1955, as interest in the device reached fever pitch, he had a different set of pressures to deal with. At home in Wilmington, Jackie was pregnant with their first child.
Having first mentioned children on their third date, Buck was ecstatic about the impending arrival. He would take home T-bone steaks and insist that Jackie eat the tenderloin. “Growing babies need good building blocks,” he repeatedly reminded her, “quality amino acids, protein.”
The summer of 1955 was a brutally hot one in Massachusetts. Jackie, weighed down with the baby, was struggling to cope. Buck, in an attempt to distract Jackie, came up with a bizarre idea.
He took home a soldering iron and a Heathkit amplifier that would become the center of a new sound system for their house. Jackie, who had never made any pretense at being technically minded, soon found herself being taught how to make perfect electrical joint connections. The amplifier she built while eight months pregnant worked flawlessly for years to come.
When the baby was due, Buck refused to leave the hospital until the baby had been born safely, subsisting on packets of M&Ms from the vending machine. The idea of a father coming into the delivery room was still very much taboo, but Buck hovered by the door for hours.
Eventually, at 7:42 p.m. on September 4, 1955, Carolyn Buck was born, weighing nine pounds and five ounces. Buck could hardly contain himself. In the days before the baby was allowed home, he bounced in and out of the hospital three times a day: morning, noon, and night. He would gaze adoringly through the nursery window. The nurses were so charmed by his doting behavior that they let him stay long after visiting hours were over.
On the day they took Carolyn home, Jackie’s mother and sister Gwen were waiting for them in the doorway. Buck took the baby from her mother’s arms and held her up for the waiting neighbors to see.
“It’s a boy next year,” he proclaimed.
His mother-in-law’s jaw dropped. “Oh my God, did you hear what he said?”