5
The Stars Are Right

In late 1967, electrical engineer Ed de Castro stood before the DEC Operations Committee, the company’s main decision-making body, and attempted to sell its members on a radical new product idea. At the time, DEC found itself in a delicate period of transition. In 1964, the company had launched its first high-end computer system, the PDP-6, a $300,000 machine optimized for time-sharing. Designed by Gordon Bell and championed by company co-founder Harlan Anderson, the machine proved too ambitious for the small company and became an expensive flop due to a faulty architecture.¹

In response, CEO Ken Olsen reorganized DEC around product lines, with the manager of each company product completely responsible for its development and market positioning and forced to bid for resources from the functional units like sales and manufacturing that comprised the Operations Committee that now had final say over which products would be built. In effect, each project manager became an entrepreneur, and the functional managers became investors allocating their resources based on which projects the Operations Committee felt deserved the strongest backing.²

Harlan Anderson resisted the new structure, particularly after Olsen put him in charge of the PDP-6 that was now clearly destined to be a failure. This conflict drove Anderson out of the company in 1966. Meanwhile, Olsen turned his back on large computers and poured resources into another computer designed by Bell called the PDP-5. Conceived at the request of a client as a front-end for another failed large computer, the PDP-4, Bell chose to make it a general-purpose computer instead, albeit one with limited functionality due to its minimalist design. DEC initially projected sales of only ten units, but at $27,000 the PDP-5 was the cheapest core-memory computer on the market by far and consequently sold roughly 1,000 units.³

¹ Rifkin and Harrar, 1988, 55–56.

² Ibid, 57–58.

The PDP-5 was one of the earliest products in a new category called the minicomputer.⁴ Unlike the hulking mainframes of IBM and its competitors – by now contemptuously referred to as the “seven dwarfs” in recognition of IBM’s total market dominance⁵ – a minicomputer was a relatively small and cheap computer that compensated for a lack of raw processing power or expansive memory through efficient design and clever tricks in order to perform most routine operations just as well as its larger cousins. Faced with the success of the PDP-5 and the failure of the PDP-6, Ken Olsen decided the minicomputer was the future of DEC.

In 1965, DEC released a follow-up to the PDP-5 designed by Ed de Castro called the PDP-8. Despite a smaller memory capacity than a full-sized mainframe, the PDP-8 was cleverly designed to perform virtually any operation a larger computer could, albeit sometimes much slower. While complex calculations might take a long time, however, many simpler operations could be performed just as quickly on a PDP-8 as on a much larger and more expensive computer.⁶ The PDP-8 was also incredibly small, as de Castro employed an especially efficient board design that allowed the entire computer to fit into a case that occupied only eight cubic feet of volume, meaning it was small enough to place on top of a standard workbench. DEC priced the computer at just $18,000 and sold 1,400 of them by 1968 to cement the minicomputer as an important new product category.⁷

Despite the success of the PDP-8, DEC found itself already behind the curve by the time it was released. In April 1964, IBM announced its System 360 series of mainframe computers, a revolutionary concept that unified the company’s entire product line from the extreme low end all the way up to its most powerful supercomputers. Previously, an expanding company that required additional computing power would not only need to acquire a new mainframe, but also need to buy new peripherals like printers and storage devices and commission new software. With the scalable 360 architecture, however, a company could upgrade to a more powerful member of the 360 family while leaving its devices and programs intact. By 1970, IBM had achieved an install base of over 35,000 System 360 computers and cemented an ironclad grip on between 70% and 80% of the mainframe computer market.⁸

³ Ibid, 59.

⁴ Ibid. The PDP-5 helped popularized the minicomputer concept, but it was not the first one designed. Earlier prototypical minicomputers included the LINC from TX-0 designer Wes Clark and the CDC 160 from brilliant computer designer Seymour Cray.

⁵ These companies, which never controlled more than 20%–30% of the mainframe market between them, were General Electric, RCA, Control Data Corporation (CDC), Burroughs, National Cash Register (NCR), Honeywell, and Sperry Rand – formed when UNIVAC maker Remington Rand merged with Sperry Gyroscope in 1955.

⁶ Ceruzzi, 2003, chap. 4.

⁷ Lamont Wood, Datapoint: The Lost Story of the Texans Who Invented the Personal Computer Revolution (Austin: Hugo House Publishers, 2012), kindle, chap. 5. By the time it was discontinued in 1990, the PDP-8 reached sales of 50,000 units. Rifkin, 1988, 71.

⁸ Ceruzzi, 2003, chap. 5.

In addition to introducing widespread compatibility, the System 360 adopted a memory word length based on multiples of eight bits rather than the previous standard of six. With IBM’s commanding lead in the marketplace, the 8-bit byte became the standard building block of computer memory – as it remains in the present day – so the PDP-8, built around a 12-bit architecture, appeared primitive by comparison. De Castro knew that his computer would soon be obsolete as the fledging minicomputer industry turned to 16-bit designs, and he looked to IBM for inspiration in crafting its successor.⁹

Now after months of work with colleagues Henry Burkhardt and Dick Sogge, de Castro proudly unveiled his minicomputer of the future to the Operations Committee, the PDP-X, which like the 360 would be based on the 8-bit byte and feature scalability across 16- and 32-bit computer architectures while retaining software compatibility. De Castro felt the PDP-X line would secure DEC’s place in the minicomputer market for decades to come, but the project proved too ambitious for Ken Olsen. The company founder was still smarting from the failure of the PDP-6 mainframe in 1964, so he was nonplussed when de Castro indicated the PDP-X project could be even more complicated and expensive. Furthermore, the managers in charge of DEC’s existing computer lines did not appreciate the thought of their products being rendered obsolete instantly, while the sales and marketing departments were wary of cannibalizing sales of the existing product line. In the end, the Operations Committee turned de Castro’s proposal down.¹⁰

Furious at what he saw as a complete lack of vision, de Castro left DEC and in conjunction with Burkhardt, Sogge, and a Fairchild Semiconductor salesman named Herb Richman established a new computer company called Data General in early 1968. Deciding that a computer family in the vein of the PDP-X would be too ambitious for a young startup, de Castro resolved instead to bring a low-cost 16-bit computer to market. The result was the Nova, a particularly small minicomputer that fit comfortably on a desktop and sold for a mere $3,950.¹¹

⁹ Rifkin and Harrar, 1988, 90–91.

¹⁰ Ibid, 87–93.

¹¹ This was just the price of the basic computer. A standard configuration with 4K of memory and a teletype cost a still very reasonable $7,950. Dag Spicer, “The Data General Nova,” Core, March 2001, 2–3.

The technological advance that paved the way for the development of a small, relatively cheap computer like the Nova was the integrated circuit (IC), in which all the components of a circuit – transistors, resistors, capacitors, etc. – are etched on a single sliver of material, commonly referred to as a “chip,” and are therefore smaller, cheaper, and more durable than discrete components connected by wires. The development of the IC was spurred by the needs of the Pentagon’s Minuteman guided missile program and the exceptionally small and durable electronic components it required. Jack Kilby at transistor market leader Texas Instruments developed the first IC in late 1958, but his design could not be mass-produced because even though he succeeded in placing all the components on a single chip he failed to eliminate wires entirely.¹²

The company that perfected the IC was one of TI’s main rivals in transistors, Fairchild Semiconductor. Founded on September 19, 1957, Fairchild was formed by eight scientists and engineers originally recruited by transistor co-inventor William Shockley for his Palo Alto-based Shockley Semiconductor, which he established in 1955 after leaving Bell Labs.¹³ The eight men had grown tired of Shockley’s jealous and controlling behavior and resolved to resign en masse and convince a company to form a research group around them so they could continue to work together. Their plans changed after they wrote to a New York investment bank, where bankers Bud Coyle and Arthur Rock advised them to think bigger and find a business willing to form an entire semiconductor company around them instead. After several companies rejected the idea, Coyle convinced inventor Sherman Fairchild that his Fairchild Camera and Instrument should put forth the capital, leading to the creation of Fairchild Semiconductor.¹⁴ The new company quickly established itself by developing a new transistor fabrication method dubbed the planar process that rendered all previous methods of transistor creation obsolete.¹⁵

¹² Young, 1998, 123.

¹³ The eight men were chemist Gordon Moore, mechanical engineers Eugene Kleiner and Julius Blank, metallurgist Sheldon Roberts, and physicists Jay Last, Jean Hoerni, Vic Grinich, and Robert Noyce.

¹⁴ Sherman Fairchild established his first company, the Fairchild Aerial Camera Corporation, in 1920 after inventing an aerial photography camera for the U.S. Army during World War I. After founding several more companies, he consolidated his businesses under the holding company Fairchild Aviation in 1927, which he renamed Fairchild Camera and Instrument in 1944 after spinning back out his aviation business.

¹⁵ Michael S. Malone, The Intel Trinity: How Robert Noyce, Gordon Moore, and Andy Grove Built the World’s Most Important Company (New York: Harper Business, 2014), 7–16.

Early in 1959, Fairchild Semiconductor co-founder Robert Noyce designed an IC that could be manufactured using the planar process, thus paving the way for mass production. In 1962, the engineers at MIT building the computer for use in the Apollo moon-landing program chose to use ICs in their architecture, and by 1964, the Apollo program’s high-volume purchases helped drive the cost of ICs down far enough that they could be used in other applications. Early ICs only integrated between one and ten transistors on a single chip, but the Nova was an early adopter of medium-scale integration (MSI) circuits, which could integrate up to 500 transistors and allowed Data General to undercut all its competitors in the minicomputer space on price and smash through the $10,000 barrier.

In December 1968, Data General unveiled the Nova at the Fall Joint Computer Conference in San Francisco in advance of its commercial release in early 1969. For this occasion, the engineers at the company hooked up a 4K Nova to a display and showcased the capabilities of the system through a version of what by now had become one of the principle programs for demonstrating the power of a small computer: Spacewar!.¹⁶ While the Nova was still too expensive to classify as a consumer electronic device, it was potentially cheap enough to incorporate into a commercial entertainment product exhibited in a public venue like a bar or bowling alley. Doing so, however, would require a visionary entrepreneur who was both comfortable with computer technology and understood the dynamics of the existing commercial amusement industry. One man who fit this profile perfectly was Nolan Bushnell.

***

Nolan Kay Bushnell was born on February 5, 1943, in Ogden, Utah, and grew up in nearby Clearfield, the home of his parents Clarence, a cement contractor, and Delma, a teacher and librarian. He discovered his true calling in the third grade when his teacher assigned him to teach a unit on electricity. Fascinated by the batteries and magnets at the school, he soon began looking for other outlets to feed his insatiable curiosity. While neither of his parents were particularly tech savvy, one of their neighbors ran an electronics surplus store and was also a HAM radio operator. The neighbor became young Nolan’s mentor, leading him to become one of the youngest HAM radio operators in the state at the tender age of 11. From there it was a small step to raiding local area junkyards for discarded components from the military bases in the Clearfield area and cultivating an interest in model rocketry.¹⁷ By the time he was a teenager, Bushnell began subsidizing his electronics habit through a job with a local firm owned by a second cousin called Barlow Electronics as a deliveryman and appliance and TV repairman.¹⁸

¹⁶ Monnens and Goldberg, 2015.

¹⁷ Nolan Bushnell, interview with the author, January 21, 2015.

¹⁸ Nolan Bushnell, Deposition, Magnavox Co. v. Bally Manufacturing Corp., No. 74-1030 (N.D. Ill), January 13–14, 1976, 13–14.

In high school, Nolan’s love of electronics cooled off a bit. He traded his HAM radio in for skis, joined the basketball and debate teams, and started reading philosophy as a hobby.¹⁹ In the summer of 1958, the Bushnell family received a shock when patriarch Clarence suffered a heart attack on the job and died. Raised in the Mormon faith – though he ultimately left it behind – Nolan had been steeped in the importance of family and hard work all his life, so at just 15 years of age he finished his father’s outstanding contracts himself. For Nolan, this feat stood as proof that with a little hard work and determination, he could accomplish anything.²⁰

In 1961, Nolan matriculated to Utah State University to study engineering, but he quickly found that the major did not agree with him. As a member of the Pi Kappa Alpha fraternity, he enjoyed partying with his fraternity brothers more than spending hours doing homework, and he soon switched his major to business.²¹ To help pay his way, he continued to work at Barlow for a time and also took several other jobs over the next few years including assembler and test tech at Litton Guidance and Control Systems, draftsman for a faculty member in the industrial engineering department of Utah State planning irrigation systems, salesman for a company called Hadley Clothing, and vendor for Encyclopedia Americana.²² His most successful venture was an advertising business he established called the Campus Company that produced a blotter three times a year containing a calendar of events for the school surrounded by advertising.²³ Worried he would fritter away his earnings through his partying lifestyle if he had nothing to occupy his evenings, Nolan accepted a new job in 1963 with the Lagoon Amusement Park in Farmington, Utah.²⁴

¹⁹ Robert Slater, Portraits in Silicon, rev. ed. (Cambridge, MA: MIT Press, 1989), 298–299.

²⁰ Steven Kent, The Ultimate History of Video Games: From Pong to Pokémon and Beyond – The Story Behind the Craze That Touched Our Lives and Changed the World (New York: Three Rivers Press, 2001), 28.

²¹ Bushnell, 2015.

²² Bushnell, January 1976, 5–8.

²³ Slater, 1989, 299.

²⁴ Bushnell, January 1976, 8.

Lagoon proved the perfect match for Nolan’s boundless energy and enthusiasm, and he soon gained a reputation as a fantastic barker and showman. Starting on the midway running a bottle toss booth, Nolan rotated through a series of carnival games over the next two years before being placed in charge of the amusement park arcade, for which he shared full profit and loss responsibility in conjunction with another employee.²⁵ This job introduced Nolan to the world of coin-operated amusements, electromechanical contraptions like pinball and shooting galleries in which playtime is regulated by inserting a coin into a slot mounted on the machine. Throughout his life, Nolan had bounced between multiple activities, academic disciplines, and vocations in his unquenchable thirst for knowledge and new experiences, but in coin-operated games, he finally found a field that satisfied his creative, technical, and entrepreneurial drives equally.

The increased time commitment of his new job rendered the hour plus commute from Utah State to Lagoon unpalatable, so in early 1965 Nolan transferred to the University of Utah to major in economics.²⁶ After marrying Paula Nielson in 1966, however, he developed a new sense of responsibility, swore of partying, and switched his major back to engineering. The same year that Nolan arrived at Utah, the school established a computer science department under the auspices of noted alumnus David Evans, who had spent the early part of the decade doing pioneering work in time-sharing and computer graphics at the University of California at Berkeley. Nolan found himself drawn to the new Utah Computer Center, and when he switched his major back to electrical engineering, he chose a focus in computer design.²⁷ He also took introductory programming courses in FORTRAN and Algol,²⁸ but his focus was on hardware rather than software.²⁹ Anxious to leave Utah, Nolan travelled to Northern California shortly before graduating in December 1968 to look for work among the high concentration of technology companies in a region that would soon be christened “Silicon Valley” due to the large number of semiconductor manufacturers in the area.

²⁵ Ibid, 9–14.

²⁶ Bushnell, 2015.

²⁷ Nolan Bushnell, Deposition, Magnavox Co. v. Bally Manufacturing Corp., No. 74-1030 (N.D. Ill), July 3, 1974.

²⁸ Bushnell, January 1976.

²⁹ Reports that he actually programmed games at the University of Utah appear to be apocryphal, as in his deposition in the Magnavox case Bushnell explicitly stated that he was not aware of any games at the institution other than Spacewar!. Nolan Bushnell, Deposition, Magnavox Co. v. Bally Manufacturing Corp., No. 74-1030 (N.D. Ill), March 4, 1976, 323.

Bushnell placed his resume in the hands of as many Bay Area tech companies as he could during his brief visit to the region and attracted his most enthusiastic response from the Ampex Corporation,³⁰ one of the world’s foremost purveyors of magnetic tape recording technology.³¹ Ampex hired him into its division developing a digital replacement for vertical filing called the Videofile Information System in which individual documents were filmed on videotape and indexed electronically for quick recall at remote readers connected to the central database by microwave links. Bushnell joined the project in early 1969 as an associate engineer to work on developing error correction solutions to increase the accuracy of the recording process.³²

Growing up, Nolan was an avid player of games, and at the University of Utah, he played number two board on the chess club. One day, Nolan opined to the number one board, a Korean, that chess was surely the most complicated game ever created. His friend disagreed and introduced Nolan to the ancient Chinese game of Go.³³ Nolan developed an obsession for the game, so when he relocated to California he began frequenting several Go clubs in the area, including one that met on the campus of Stanford University. At the Stanford club, Nolan became friendly with a man named Jim Stein who worked at the Stanford AI Lab. One night, Jim offered to take Nolan over to the lab to show him the cool programs they had running on their computer,³⁴ thus introducing Nolan to one of the favorite pastimes of the AI Lab students and employees: Spacewar!³⁵

³⁰ Bushnell, 2015.

³¹ Established as a motor manufacturer by Alexander Poniatoff in 1944, Ampex pivoted to recording technology after World War II, releasing the first magnetic tape recorder available in the United States in 1948 and the world’s first videotape recorder in 1956.

³² Morgan Ramsey, Gamers at Work: Stories Behind the Games People Play (New York: Apress, 2012), Kindle, chap. 2.

³³ Bushnell, 2015.

³⁴ Ramsay, 2012, chap. 2.

³⁵ Mr Bushnell has, of course, claimed for decades that he first saw and played Spacewar! as a student at the University of Utah in the mid-1960s. The earliest known instance of this claim dates from a November 1973 article in Systems Engineering Today, and Bushnell testified to the same while under oath during multiple depositions in the mid-1970s. There are serious reasons to doubt this claim due to a lack of corroborating evidence that Spacewar! was present at Utah before the early 1970s. In-depth research in university records of the computer equipment available at Utah in the 1960s by Marty Goldberg indicated no machines capable of playing the game were present, other students affiliated with the computer center at the time do not remember the game (Randall Willie, interview with the author, May 25, 2017), and the university newspaper makes no mention of it, unlike at MIT and Stanford where it was covered soon after appearing. Furthermore, in the earliest known interview with Bushnell, conducted for a documentary called Games Computers Play filmed in early 1973, Nolan explicitly states he was inspired to create a commercial video game due to his experience playing games at Stanford and does not mention Utah at all. The narrator of the documentary then states that Bushnell first saw Spacewar! at Stanford.

From the moment Nolan arrived at Ampex, he had been plotting his departure. Restless and fiercely independent, he was only truly happy when he was operating on his own dreaming up the next big thing in technology and/or entertainment. Indeed, the place he had really hoped to work was the Walt Disney Company, which employed a group of engineers dubbed the Imagineers to develop new attractions for its theme parks, but the company did not hire fresh graduates.³⁶ After becoming enthralled by Spacewar!, Nolan felt he should commercialize the game, but like so many players before him he could not initially figure out how to do so. That changed in spring 1970 when he happened upon a sales flyer for the Nova minicomputer. With his arcade background, Bushnell realized that at $4,000 a Nova cost the same as roughly a half dozen standard electromechanical arcade games. Therefore, if he could figure out a way to time-share the computer so that six users could play simultaneously and scrounge up some cheap displays, he just might have a viable commercial product.

***

Nolan Bushnell was creative, energetic, even visionary, but he was not a particularly accomplished engineer. Although he remained an eager learner and read up constantly at Ampex, too many years of partying, running side businesses, and periodically switching majors left him without sufficient background to handle an ambitious project like a commercial computer game by himself. Fortunately, he shared his office with an older and more experienced engineer who possessed many of the skills he lacked named Ted Dabney. Born in 1937 in San Francisco, California, Samuel Frederick “Ted” Dabney was the complete opposite of Bushnell. Whereas Bushnell was creative, Dabney was stodgy. Whereas Bushnell was a dreamer, Dabney remained focused on day-to-day practicalities. And whereas Bushnell was ambitious, driven, and entrepreneurial, Dabney was perfectly content working for a large company on whatever projects came his way.

³⁶ Slater, 1989, 303.

As a youth, Dabney had been aimless and barely completed high school, though he did display an aptitude for mathematics. His geometry skills earned him employment as a surveyor, but work proved hard to come by in this seasonal profession. With college out of the question, Dabney enlisted in the U.S. Marine Corps and received a thorough grounding in electronics through a 16-week course at Naval Station Treasure Island in San Francisco supplemented by a radio relay course at the Marine Corps Communication Electronics School in San Diego. After leaving the Corps in 1959, Dabney worked briefly for Bank of America and Hewlett-Packard before joining the Military Products Division of Ampex. One of his primary projects there was creating video amplifiers and gamma correctors for an electron beam scanning system intended to digitally convert U-2 spy plane footage, so when the Videofile project commenced in 1966, Dabney was one of the first engineers brought on board. For Videofile, Dabney’s primary duties were adapting a vidicon camera for use with the system, evaluating monitors and building the circuitry to allow them to interface with the system, and designing additional components such as power supplies. Much of the circuit design Dabney contributed to the project was virtually identical to the work he did in military products except that he used transistors rather than vacuum tubes.³⁷

When Bushnell joined the Videofile team three years later, he and Dabney bonded over their similar family lives – despite their age differences, they both had daughters roughly the same age – and a shared love of engineering. Bushnell taught Dabney to play Go and spent many a lunch hour engaged in matches with the older man. After deciding to transform Spacewar! into a commercial product, Bushnell took Dabney up to Stanford to see the game for himself in an effort to enlist his aid in building a commercial version of the game. When Bushnell described his plan for a time-shared coin-op system, Dabney was not quite sure what to make of it, but he was perfectly happy to give it a try.³⁸

³⁷ Samuel F. (Ted) Dabney, interview by Chris Garcia, July 16, 2012, Computer History Museum, https://archive.computerhistory.org/resources/access/text/2012/10/102746459-05-01-acc.pdf.

³⁸ Marty Goldberg and Curt Vendel, Atari, Inc.: Business is Fun (Carmel, NY: Syzygy Company Press, 2012), kindle, chap. 1.

Between them, Bushnell and Dabney possessed all the analog, digital, and video engineering skills required to build interfaces between a Nova computer and a display, but to develop the necessary time-sharing routines to run multiple instances of Spacewar! they required a programmer. For this task, Bushnell turned to another friend in the Videofile division with whom he often played chess and Go at lunch and socialized with outside of work named Larry Bryan. A mathematician from Florida, Bryan entered the programmer’s trade strictly by accident after answering an ad from UNIVAC in 1963 while waiting to take a teaching job that fall. Bryan took to his new profession quickly and bounced around several companies working on defense projects before coming to Ampex in 1967 as the first programmer in the Videofile division. Like Dabney, Bryan proved content to go along with Bushnell’s scheme, and a three-way partnership was born.³⁹

In the summer of 1970, Bushnell, Dabney, and Bryan held several meetings to finalize their plans and agreed that Bushnell would handle the electronic engineering for the game, Dabney would do the video engineering, and Bryan would program the software.⁴⁰ They also chose a name for their venture. Bushnell and Dabney originally wanted something incorporating their initials such as D&B Enterprises, but decided D&B could be confused with Dunn and Bradstreet, while B&D might be mistaken for Black & Decker.⁴¹ They were at an impasse until Bryan suggested another name, Syzygy, because they were forming a partnership of three people and he remembered that the term had something to do with the confluence of three things. The trio proceeded to look up the word in the dictionary and discovered that it referred to a straight-line configuration of three celestial bodies in a gravitational system.⁴² Satisfied with this definition, the trio adopted the name for what at this point was an informal partnership.

Around August 1970, Larry Bryan commenced work on the game and spent roughly two weeks tinkering with a program for the Nova. The results were not promising. The Nova was capable of running Spacewar! – witness the Joint Computer Conference demonstration in 1968 – but it had nowhere near enough power to run four to six versions of the game at once, which was the only way to make the entire enterprise economical. Despite his best efforts, Bryan could simply not see how to make the concept work, so after delivering the preliminary results to Bushnell, he left the partnership.⁴³

³⁹ Larry Bryan, unpublished interview by Martin Goldberg, March 2015.

⁴⁰ Ibid.

⁴¹ Goldberg and Vendel, 2012, chap. 1.

⁴² Bryan, 2015.

⁴³ Ibid.

Although Bryan declared a cost-effective Spacewar! coin-operated game impossible, Bushnell continued to work on the problem over the next several months. Realizing the whole game could not be executed in software, he decided to move as many functions as he could into specialized hardware instead, which would reduce the strain on the computer.⁴⁴ An informal lab was established in the bedroom of Ted Dabney’s daughter, where the hardware slowly came together.⁴⁵ To fund the project, Bushnell and Dabney each contributed $100, which proved more than sufficient to buy components that generally cost only $0.15–$0.25 each, especially since they could procure many of the items they needed from Ampex, which encouraged employees to take small amounts of inventory for their own projects.⁴⁶ The most expensive parts of the system were a used television purchased from Goodwill Industries and a power supply, neither of which cost more than $30.⁴⁷ As a stand-in for the computer, Bushnell and Dabney built an “exerciser,” a simple piece of hardware that emulated some of the capabilities of the Nova and allowed them to generate and move a dot around the screen to make sure the rest of the equipment was working properly.⁴⁸

Once he was satisfied with the basic hardware, Bushnell opened a dialogue with Data General in January 1971 about ordering a few Nova 1200 computers, a cheaper model of the computer introduced at the end of 1970.⁴⁹ With their venture finally ready to begin in earnest, Bushnell and Dabney also took steps that same month to formalize their business relationship by officially organizing Syzygy, Co. as a partnership. As part of this process, both partners invested another $250 into the venture, bringing their total contributions to $700.⁵⁰ On January 26, 1971, Bushnell drafted a letter to a salesman at Data General ordering “six of everything” so that he and Dabney could build their first games.⁵¹ He never sent it.

⁴⁴ Bushnell, January 1976, 40–41.

⁴⁵ For decades, Mr Bushnell claimed that the prototype unit was built in his own daughter’s bedroom, which is the version of events reported in most secondary media. In 2009, Ted Dabney emerged after almost 40 years of silence to tell his side of the story and claimed that it was built in his daughter’s bedroom. Since then, Bushnell has acknowledged that some of the work took place at Dabney’s house, but continues to claim work proceeded at his house as well. Based on the recollections of both Dabney and Dabney’s daughter combined with Bushnell’s partial admission, I believe Dabney’s story to be more plausible.

⁴⁶ Kent, 2001, 31.

⁴⁷ Ramsay, 2012, chap. 2.

⁴⁸ Bushnell, January 1976, 59–60.

⁴⁹ Ibid, 50–55.

⁵⁰ “Statement of Owner’s Equity: Year Ended December 31, 1971 (Unaudited),” Syzygy Company, c. 1972.

⁵¹ Bushnell, January 1976, 50–51.

At some point in late January or early February 1971, Bushnell reserved some time on a local area Nova to make sure his program would run properly on the actual computer. One of the staff at the computer facility looked over the program and called Bushnell’s attention to some errors in his calculations. Due to these problems, it turned out that what Bushnell and Dabney were now calling Cosmic Combat would not run smoothly on the Nova even after moving much of the functionality to specialized hardware.⁵² For a day or two, Bushnell was despondent, but then he realized that between the specialized hardware they had constructed and the exerciser that replicated some of the Nova’s functionality, he and Dabney had practically created the entire game in hardware already, and it would be relatively simple to increase the complexity of the exerciser and do away with the computer altogether.

By spring 1971, Nolan and Ted had completed a rough prototype of Cosmic Combat. With their hardware plans now entering their final stage, the duo turned their attention to securing the financial backing necessary to produce the game. They first tried to interest Ampex in the project through their boss, Ed De Benedetti, but he turned them down. Next, Dabney turned to one of his mentors, a former engineer in the Ampex Military Products Division named Irving Roth, but he turned them down as well. Unsure where to turn next, Bushnell made a fortuitous connection through his dentist. During a routine dental appointment, Bushnell began describing the game he was working on and learned that a fellow patient named Dave Ralstin served as the sales manager of a local firm called Nutting Associates, one of a very few companies involved in the manufacture of coin-operated games not located in the industry hub of Chicago. Bushnell contacted Ralstin, and two days later delivered his pitch to create a video arcade game for the company.⁵³ By April 1971, Bushnell had resigned from Ampex and joined Nutting as chief engineer,⁵⁴ where he prepared to unleash the first commercial video game through the coin-operated amusement industry, which had played host to a wide variety of novelty entertainment concepts over the past century, none more significant than the game of pinball.

⁵² Ibid, 56–57.

⁵³ Goldberg and Vendel, 2012, chap. 1.

⁵⁴ Bushnell, January 1976, 72.