Chapter 15 - The Fairchild Effect

Cyan Engineering/Atari

Ampex Connection

An early addition to the tech community in the area was brought by a couple of Ampex Alumni out of the bay area. That was the company that first made video tape recording practicable. Even in the early 70s, the legendary rush hour gridlock in the San Francisco bay area, from San Jose northward, all the way to Santa Rosa, was in full display. Larry Emmons, an engineer for Ampex had had enough. He moved up to an area about 14 miles south of Grass Valley and started a consulting company. He was soon followed by Steve Mayer, who was another Ampex engineer. They started a company called Arvin Engineering in 1972.

They secured a contract with the Department of Defense to develop a video disk recorder. That project was soon canceled, and the two cast about for something else to do. For awhile they simply provided consulting services. Larry and Steve's boss at one time was Nolan Bushnell. Bushnell would go on to start Atari. This connection would forever link the Grass Valley area to video game lore.

At Ampex, Nolan Bushnell met fellow employee Ted Dabney and found they had common interests. Bushnell worked at a gaming arcade in college and developed a fascination with the electro-mechanical games of the time. Bushnell shared his ideas of creating pizza parlors filled with electronic games with Dabney. Bushnell took Dabney to the computing labs at Stanford Artificial Intelligence Laboratory to show him Spacewar!. The game was developed in 1962 at MIT and was written to run on a DEC PDP-1 mini-computer. That computer could have up to 64K of magnetic core memory and a clock speed of around 200K. Quaint by today’s standards.

In 1971, Nolan Bushnell and Ted Dabney founded a small engineering company, Syzygy Engineering with the intention of producing a Spacewar clone known as Computer Space, a game where the player shot at UFOs. They designed the game using their knowledge of video circuitry. It was not computer driven. While the first microprocessors were being introduced, the Intel 4004 4-bit microprocessor the year before, they were too expensive and underpowered. The use of analog video circuitry allowed the electronics to be much cheaper and smaller. Dabney built the prototype and Bushnell shopped it around, looking for a manufacturer. They made an agreement with Nutting Associates, a maker of coin-op trivia and shooting games, that produced a fiberglass cabinet for the unit that included a coin-slot mechanism.

Although sales of the game, which consisted of 1500 units totalling $3 million, it was considered a commercial failure. Bushnell laid the blame on Nutting, as Bushnell thought that they had not marketed the game very well. He decided that his next game would be licensed to a bigger manufacturer. Bushnell and Dabney worked for Nutting during that time, but he got Nutting to put a Syzygy Engineered label on each machine. Bushnell also knew that the next game they developed would need to be simpler and not require users to read instructions on the cabinet, since their target audience would likely be drunken bar patrons.

Arcade Games

They left Nutting and got funding from Bally Manufacturing to develop a new game. Bushnell and Dabney decided it was time to incorporate. Amazingly, Syzygy, an astronomical term, was already taken in the state of California. Bushnell was a fan of the board game "Go," and a term in the game was "Atari," which essentially in the context of the game meant what "check" meant in chess. On June 27, 1972, the two incorporated Atari, Inc. It actually means "to hit a target" in Japanese.

They hired their first design engineer, Al Alcorn, who was another Ampex alumni. At Cal Berkeley he participated in a work-study program that brought him to Ampex, where he worked under Bushnell. He graduated just as Bushnell was forming Atari.

After Bushnell attended a Burlingame, California demonstration of the Magnavox Odyssey, he gave the task of making the Magnavox tennis game into a coin-op version to Alcorn as a test project. He told Alcorn that he was making the game for General Electric, in order to motivate him, but in actuality he planned to simply dispose of the game, using it as a training episode for Alcorn. Bushnell's initial intention was to then move on to a car racing game, which would have been a tougher engineering design project. Bushnell thought that game would be more appealing than spaceship navigation to beer drinkers in a workingman's bar.

But the "warm up" project when completed seemed that it might have some appeal. Bushnell offered the finished design to Bally, who rejected it because the game required two players. Bushnell decided to build a prototype anyway and place it in a bar near them. That game was Pong. Atari released several sequels to Pong that built upon the originals game-play by adding new features. Not long after Pong's release Bushnell and Dabney had a falling-out. Bushnell ended up purchasing Dabney's share of Atari for $250,000 in 1973.

Atari was in the arcade gaming business, but they faced a problem with how that business operated. Local distributors of the games insisted on exclusive deals with suppliers. So in 1973, Atari secretly spawned a competitor called Kee Games, headed by Nolan's next door neighbor Joe Keenan, to sidestep pinball distributors' insistence on exclusive distribution deals. So both Atari and Kee could market virtually the same game to different distributors, each thinking they had an "exclusive" deal. Joe Keenan's management of the subsidiary led to him being promoted president of Atari that same year.

Soon after Pong's release, several companies began producing games that closely mimicked its game-play. Eventually, Atari's competitors released new types of video games that deviated from Pong's original format to varying degrees, and this, in turn, led Atari to encourage its small staff to move beyond Pong and produce more innovative games themselves. Bushnell knew that with Magnavox's Odyssey game taking off with consumers, he needed to move beyond arcade games. 'Screen Games' as Time magazine called them at the time, estimated that it was going to be a $900 Million a year business.

Bushnell discovered that his former colleagues at Cyan Engineering were looking for work, so in 1973 he offered Larry and Steve a contract to help Atari develop what eventually became known as the Atari VCS console. The company changed its name to Cyan Engineering. Interestingly it had relocated to the Litton building. Where the Grass Valley Company had gotten its start.

Bushnell knew he had to start moving in a new direction swiftly. Atari sold 8500 Pong machines and bought in $14 Million in 1974, but at the same time Pong sales were slowing. The company was growing too fast and had missteps with a couple new games. Also it opened a subsidiary in Japan that turned out to be a financial disaster. In May of 1974 Atari reported a $500,000 loss, and they laid off half the staff. Some were saying that the company looked like a one-trick pony. To help right the ship Atari and Kee merged at the end of 1974, and Joe Keenan became the president of Atari, while Bushnell kept the CEO title, Steve Bristow became head of Engineering, and Al Alcorn became head of R&D.

The company recovered the next year during the 1975 Christmas season, by releasing a home version of Pong exclusively through Sears retail stores. The home version was a commercial success. But soon it was competing with clones. Bushnell start relying heavily on Cyan Engineering.

Video Console System

By 1975, Ron Milner and Steve Mayer of Cyan Engineering had developed the first prototype of the Atari VCS. Joe Decuir was hired later that year to help with debugging the prototype and to help with building the first gate-level prototype. Decuir was a recent graduate from University of California, Berkeley. He had been interested in the 6502 microprocessor chip, which would become central to the VCS.

At this point in time Atari purchased Cyan Engineering and renamed it the Grass Valley Research Center.

The VCS was to be centered around two things, a MOS Technologies 6507 microprocessor, part of their 6502 series and an IC known as the Television Interface Adapter (TIA). The person instrumental in tying those two components together was Jay Miner. An accomplished microprocessor IC designer, he had co-founded Synertek in 1973, where he served as the company's primary chip designer. One of the company's earliest contracts were to create CMOS chips for the Bulova Watch Company. The company quickly became a second source manufacturer for chips designed by other firms such as Intel, Rockwell, and MOS Technology.

Programmable microprocessors would enable a console design significantly simpler and more powerful than any dedicated single-game unit. At the time microprocessors were expensive, $100–300 each, pricing most of them far outside the range of what the VCS market would pay.

Atari had opened negotiations to use Motorola's new 6800. But in September 1975, MOS Technology debuted the 6502 microprocessor for $25 at the Wescon trade show in San Francisco. Mayer and Milner attended and met with the leader of the team that created the chip, Chuck Peddle. Peddle proposed using the 6502 in a game console and offered to discuss it further at Cyan's facilities after the show.

Chuck Peddle was an electrical engineer, and while at Motorola recognized a market for a very low price microprocessor and began to champion such a design to complement the $300 Motorola 6800. His efforts were frustrated by Motorola management, and he was told to drop the project. He then left for MOS Technology, where he headed the design of the 6502 family of processors. Peddle was best known as the main designer of the MOS Technology 6502 microprocessor. It was introduced in 1975, and was priced at 15% of the cost of an Intel 8080. It was the 6502 that was subsequently used the Apple II, Commodore PET and Commodore VIC-20.

The MOS 6502 used a new manufacturing process that produced a much higher yield rate than previous methods, which generally saw a low percentage of ICs that worked. The 6502 was so cheap that many people believed it was a scam when MOS first showed it at a 1975 trade show. Soon both Intel and Motorola were forced to dramatically lower their prices.

At about the same time the 6502 was being released, MOS's entire calculator IC market collapsed, and its prior existing products stopped shipping. Soon they were in serious financial trouble. Another company, Commodore Business Machines, had invested heavily in the calculator market and was also nearly wiped out by TI's entry into the market.

A fresh injection of capital saved Commodore, and allowed it to invest in company suppliers in order to help ensure their IC supply would not be upset in this fashion again. In late 1976, Commodore, publicly traded on the NYSE with a market capitalization around US$60 million, purchased MOS Technologies (whose market cap was around US$12 million) in an all-stock deal.

Back at Cyan, over the course of two days, MOS and Cyan engineers sketched out a 6502-based console design based on Meyer and Milner's specifications. Financial models showed that even at $25, the 6502 would be too expensive. Peddle then offered the 6507 microprocessor, a cost-reduced version of the 6502. Cyan and MOS negotiated the 6507 and its accompanying IO chip at $12 a pair.

MOS introduced Cyan to Microcomputer Associates, who had separately developed debugging software and hardware for MOS microprocessors, and had developed the JOLT Computer for testing the 6502. Jolt was the first 6502 single board computer. Jolt sold the board as both a kit for $249, or fully assembled and tested for $348.

Milner was able to demonstrate a proof-of-concept for a programmable console by implementing Tank, an arcade game by Atari's subsidiary Kee Games, on the JOLT.

Atari wanted a second source of the chip set. Peddle suggested Synertek whose co-founder, Bob Schreiner, was a friend of Peddle. In October 1975, Atari informed the market that it was moving forward with MOS.

Due to its manufacturing of the MOS Technology 65xx series of chips, Synertek was recommended as a partner to Atari, Inc. after it had been decided to use the MOS 6507 for the VCS console. One of Atari's engineers, Harold Lee, had worked with Miner at Standard MicroSystems and suggested him as the designer for a custom chip which would power Atari's new console. Through an arrangement with Synertek, Atari hired Miner in late 1975 to lead the chip design for the Atari VCS, primarily that of the display hardware, the TIA. The TIA sent graphics and audio to a television through via RF into a TV sets antenna connector.

By March 1976 with the help of Miner, the group in Grass Valley had a prototype that included included a TIA, a 6507 microprocessor, and a ROM cartridge slot and adapter. Decuir who was working with Miner on the design had apprenticed for Miner.

Atari also sent up Ted Hoth, who had been involved with the developing the Intel 4004, the very first microprocessor. Ted Hoth was a imposing guy at 6'4", who also had a background in nuclear metallurgy. Also on the team was hire Jim Heb, George Foster, John Sperry, and Herb Foster, who eventually started Viking Video, which made video disc recorders. Ironically back to Cyan's roots. Another young engineer working on the project was Chuck Meyer, today still a prominent player with the Grass Valley Company. He was doing doing development in 6502 assembly language for the 6507 the VCS was using.

The 6502 MOS Technologies salesman was aggressive and generous in handing out samples as the microprocessor business had become very competitive at that time. He seemed to have had a jewel case with "chemicals" he gave away along with the chips. That was when the 8-bit Intel 8080 was $1000 each.

While the base 6502 microprocessor was a more capable IC than the 6507 that was decided on, the 6507 was much cheaper. There were significant limitations in the 6507, which would run at 1.19 MHz in the VCS, which would be named the 2600, which while not lightning fast, it was fast enough. The 6507 was cheaper than the 6502 because its package included fewer memory address pins—13 instead of 16, reducing the already limited addressable memory to 4 KB. This was believed to be sufficient as Combat is itself only 2 KB. Later games circumvented this limitation with memory bank switching. This is where blocks of memory are switched between as needed.

As the VCS design became more refined, in 1977 Al Alcorn brought in Atari's game developers to provide input on features, as they had to modify their games to be ported to the VCS.

Never-the-less Atari was beat to market in November 1976 by The Fairchild Channel F, short for "Channel Fun." It was the first video game console to use ROM cartridges, instead of having games built-in, and the first console to use a microprocessor. It had a retail price of US$169.95. Equivalent to $772.92 in 2020.

Bushnell brought in Gene Landrum, a consultant who had consulted for Fairchild on its Channel F, to determine the consumer requirements for the console. Landrum suggested the following:
• A living room aesthetic, with a wood grain finish
• The cartridges must be "idiot proof, child proof and effective in resisting potential static [electricity] problems in a living room environment."
• Recommended it include four to five dedicated games in addition to the cartridges, which was ignored.
• The cartridge design was done by James Asher and Douglas Hardy. Hardy had been an engineer for Fairchild and helped in the initial design of the Channel F cartridges, but he quit to join Atari in 1976.

The interior of the cartridge that Asher and Hardy designed was sufficiently different to avoid patent conflicts, but the exterior components were directly influenced by the Channel F to help work around the static electricity concerns. Atari encountered production problems during its first batch, and its testing was complicated by the use of cartridges.

Warner

Bushnell knew he had another potential hit on his hands but bringing the machine to market would be extremely expensive. Atari was still recovering from its 1974 financial woes and needed additional capital to fully enter the home console market, though Bushnell was wary of being beholden to outside financial sources. Atari obtained smaller investments through 1975, but not at the scale it needed, and began considering a sale to a larger firm by early 1976. Looking for outside investors, Bushnell sold Atari to Warner Communications in October 1976 for an estimated $28–32 million. Atari revenue at that time was around $30 million/year.

Warner saw the potential for the growing video game industry to help offset declining profits from its film and music divisions. Negotiations took place during 1976, during which Atari had to clear itself of some liabilities. Those included settling a patent infringement lawsuit with Magnavox over Ralph H. Baer's patents that were the basis for the Magnavox Odyssey.

Ralph Henry Baer was a German-American inventor, game developer, and engineer. Baer's family fled Germany just before World War II, and Baer served the American war effort, gaining an interest in electronics. After several jobs in the electronics industry, he was working as an engineer at Sanders Associates (now BAE Systems) in Nashua, NH. There, around 1966, he conceived the idea of playing games on a television screen. With support of his employers, he worked through several prototypes until he arrived at a "Brown Box." That box would later become the blueprint for the first home video game console, licensed by Magnavox as the Magnavox Odyssey.

Warner provided an additional $120 million which allowed the project to be fast-tracked. The Atari VCS console was finally released in 1977. The unit was showcased on June 4, 1977 at the Summer Consumer Electronics Show with plans for retail release in October. The announcement was purportedly delayed to wait out the terms of the Magnavox patent lawsuit settlement, which would have given Magnavox all technical information on any of Atari's products announced between June 1, 1976 and June 1, 1977. The introductory price of $199 (equivalent to $894 in 2019) included a console, two joysticks, a pair of paddles, and the Combat game cartridge.

Warner was a traditional, proper normal business. In Silicon Valley, Atari was not normal business. It was a very hit-driven technology business. It was an industry where you had to eat your own young because if you didn't, somebody else would. Warner management team did not understand this. Instead they ran Atari like a big, bureaucratic organization that supported it current products.

Atari created unique products and had played a big part in started a whole new industry. Sony's founder Akio Morita, once came to Silicon Valley to see Atari. He said they don't plan; they just build it.

Bushnell's exit

In 1978, Atari sold only 550,000 of the 800,000 systems manufactured. This required further financial support from Warner to cover losses. After a couple of years Warner decided that Atari had to run like a normal business, so Warner installed their professional managers, who came out of traditional management where they're used to having products that don't have short life cycles.

Al Alcorn stated that "When companies take over another company, they often fail, because these fat, gray-haired men on the board don't understand culture, and they just see the business relationship. They try to merge two companies together and if the cultures don't work, it's never going to work. And this happens time and time again. It seems great on the spreadsheet or paper." Warner had bought Atari and treated it as a cash cow. There were no long-term plans.

In February 1978 Ray Kassar was hired as president of Atari Inc.'s consumer division by Warner Communications. He had previously been executive vice-president of Burlington Industries, the world's largest textile company at the time, and president of its Burlington House division. A member of the Board of Directors, Kassar had spent over thirty years at Burlington.

During the Kassar years, Atari Inc.'s sales grew from $75 million in 1977 to over $2.2 billion just three years later. Though Atari enjoyed some of its greatest success during this period, the stifling atmosphere and lack of royalties or recognition to the individual game designers angered employees, many of whom quit. During this period, nearly all members of the original Atari Inc. staff, including Al Alcorn, quit, or were fired. Rifts had begun to develop between the original Atari Inc. staff, most of whom had engineering backgrounds, and the new hires brought in by Warner, who, like Kassar, mostly had business backgrounds. Atari Inc. upper management also suffered severe turnover rates. Many blamed Ray Kassar's autocratic management style.

In December 1978 Nolan Bushnell was fired following an argument with Manny Gerard. Bushnell said that Warner claim they fired me, I say I quit. It was a mutual separation. Gerard was a member of the Office of the President for Warner during the Atari days from 1976, when they acquired Atari, to 1984 when it sold to the Atari. He was a key person in the decision for Warner to acquire Atari.

Bushnell had concerns on Kassar's plans and feared they would produced too many units to be sold, and at a board meeting with Warner near the end of the year, reiterated this position. Bushnell's concerns never materialized as a combination of Kassar's marketing and the popularity of game licensed from a Japanese company, Space Invaders drove Atari VCS sales. Both Warner Communications and Bushnell commonly recognized he was no longer a good leader for the company, removing him as CEO and Chairman in early 1979.

Warner did offer Bushnell the opportunity to stay as a director and creative consultant, but Bushnell refused. Before leaving, Bushnell negotiated the rights to Pizza Time Theater from Atari for $500,000. It had been created by Bushnell, originally as a place where kids could go and eat pizza and play video games, which would therefore function as a distribution channel for Atari games. Chuck E. Cheese's Pizza Time Theater also had animatronic animals that played music as entertainment. It was known that Bushnell had always wanted to work for Walt Disney but was continually turned down for employment when he was first starting out after graduation; Chuck E. Cheese was his homage to Disney and the technology developed there. In 1981 Bushnell turned over day-to-day food operations of Chuck E. Cheese's to a newly hired restaurant executive and focused on Catalyst Technologies.

Keenan replaced Bushnell but left a few months later, with Kassar being named as Atari's CEO by mid-1979

Atari sold more than 1 million consoles over the 1979 holiday season, but there was new competition from the Mattel Electronics Intellivision and Magnavox Odyssey, which also used swappable ROM cartridges.

At one point four game programmers wanted a small commission, but when they asked Kassar about that, David Crane recalls that Kassar responded, "You are no more important to that game than the guy on the assembly line who puts it together." So Crane and three others resigned from Atari and formed their own company: Activision, which became the first ever third party game developer. The new company formed by Crane, Whitehead, and Miller in 1979, started developing third-party VCS games using their knowledge of VCS design and programming tricks, and began releasing games in 1980.

In 1980, Atari attempted to block the sale of the Activision cartridges, accusing the four of intellectual property infringement. The two companies settled out of court, with Activision agreeing to pay Atari a licensing fee for their games. Kaboom! (1981) and Pitfall! (1982) are among the most successful with at least one and four million copies sold, respectively.

As soon as the 2600 started shipping, Atari started development on its replacement, and halfway into the development the home computer business exploded so that replacement became a full computer, the Atari 800. It sold from 1979 to 1992. It had a MOS technology 6502. At the height of its run between 1979 and 1985 the 800 sold 2 million units.

In 1980 the Commodore 64 debuts. At the time Guinness world records claims it was the highest selling single computer model of all time. At its height it was selling 2 million units per year. A number of months they were building 400,000 64s a month.

In addition to third-party game development, Atari also received the first major threat to its hardware dominance from "Colecovision." Coleco had a license from Nintendo to develop a version of the smash hit arcade game Donkey Kong (1981), which was bundled with every Colecovision console. Coleco sales then doubled again for the next two years; by 1982, 10 million consoles had been sold, while its best-selling game was Pac-Man at over 7.9 million copies sold by 1983. Coleco gained a 17% share of the hardware market in 1982 compared to Atari's 58%. With third parties competing for market share also increasingly competing for market share.

In 1982, the 2600 cost Atari only about $40 to make and was sold for an average of $125, equivalent to $340 in 2020. In 1982, Atari launched its second programmable console, the Atari 5200. The company spent $4.50 to $6 to manufacture each cartridge, plus $1 to $2 for advertising, wholesaling for $18.95, equivalent to $50 in 2020. Atari worked to maintain dominance in the market by acquiring licenses for popular arcade games and other properties to make games from.

Video Game Crash

Heading into the 1982 holiday shopping season, Atari had placed high sales expectations on "E.T. the Extra-Terrestrial," a game programmed in about six weeks. Atari produced an estimated four million cartridges, but the game was poorly reviewed, and only about 1.5 million units were sold.

Warner Communications reported weaker results than expected in December 1982 to its shareholders, having expected a 50% year-to-year growth but only obtaining 10–15% due to declining sales at Atari. Atari's major competition at the time was the Commodore 64. Coupled with the over-saturated home game market, Atari's weakened position led investors to start pulling funds out of video games, beginning a cascade of disastrous effects known as the video game crash of 1983.

Many of the third-party developers formed prior to 1983 were closed. In September 1983, Atari sent 14 truckloads of unsold Atari 2600 cartridges and other equipment to a landfill in the New Mexico desert, later labeled the "Atari video game burial." Long considered an urban legend that claimed the burial contained millions of unsold cartridges, the site was excavated in 2014, confirming reports from former Atari executives that only about 700,000 cartridges had actually been buried.

Atari reported a $536 million loss for 1983 as a whole. In July 1983, Kassar was fired due to continuing massive losses at Atari. In December 1982, Kassar had sold 5,000 shares of stock in Warner Communications only 23 minutes before a much lower than expected fourth quarter earnings report would cause Warner stock to drop nearly 40% in value in the following days. The Securities and Exchange Commission accused Kassar and then Atari Inc. vice-president Dennis Groth of trading stock with illegal insider knowledge. Kassar settled, returning his profits without acknowledging guilt or innocence.

Atari had invested heavily in games for the 2600, that became commercial failures and contributed to the video game crash of 1983 which ended the market relevance of the 2600. Some of the crash is attributed to saturation of the market for video games, standalone that is, and the rise of home computers.

Atari continued to lose money into 1984, with a $425 million loss reported in the second quarter. Coleco's success with Donkey Kong greatly hurt Atari that year. By mid-1984, third party software development for the 2600 had essentially stopped except by Atari and Activision. By 1984 games for the 2600 were using more than four times the ROM of the initially launch games.

Fire Sale

Warner, wary of supporting its failing Atari division, started looking for buyers in 1984. Warner sold most of Atari to Jack Tramiel, the founder of Commodore International, in July 1984 for about $240 million. Interestingly it was Tramiel who rescued MOS Technologies, preserving the microprocessor that his computers, and that Atari's VCS ran on. Warner retained Atari's arcade business and Ataritel, who we will look at shortly.

Tramiel was a proponent of personal computers, and de-prioritized further Atari console development following the sale. Tramiel's offer to Atari was to do a debt swap. He said it was worth more than Atari was. Taking advantage of Warner Communication's financial situation, the neighboring company of Time Inc. announced on March 4, 1989, that the two companies were to merge

Both Mattel and Coleco left the video game market by 1985. The North American video game market did not recover until about 1986, after Nintendo's 1985 launch of the Nintendo Entertainment System in North America. Video game revenues went from 3.2 billion in 1983 and fell to around 100 million in 1985, a drop of almost 97%.

Tramiel' Atari released a redesigned model of the 5200 in 1986, supported by an ad campaign touting a price of "under 50 bucks," bt it wasn't what led the recovery of the industry, it was Nintendo. With a large library of cartridges and a low price point, the 5200 continued to sell into the late 1980s. Atari released the last batch of games in 1989–90. After more than 14 years on the market, the 5200 line was formally discontinued on January 1, 1992, along with the Atari 8-bit family of home computers.

AtariTel

While not an Atari initiated project, in 1981 the Cyan group started on a "skunk works" type project. The genesis of this were a number of court cases starting in the 50s that forced AT&T to relinquish its iron grip on its phone network.

In 1956 the court had ruled that Hush-A-Phone Corporation couls sell a small, cup-like device which mounted on the speaking party's handset microphone, reducing the risk of conversations being overheard and increasing sound fidelity for the listening party. AT&T claimed that was not allowed. The court ruled differently stating that devices could mechanically connect to a phone-company-owned telephone without the permission of AT&T.

With the rise of corporate computer use in the 60s and into the 70s many customers and computer manufacturers wanted to be able to connect peripherals to main frames over phone lines. The FCC looked into a lawsuit that Carterfone had filed against AT&T. In 1968, the Federal Communications Commission extended this privilege by allowing the Carterfone and other devices to also be mechanically connected to the AT&T network, as long as they did not cause harm to the system. This device was acoustically, but not electrically, connected to the a telephone. It was electrically connected to a base station of the mobile radio system, and got its power from the base station. When someone on a two-way radio wished to speak to someone on a "landline" the radio station operator at the base would dial the requested telephone number. When the call was connected the handset of the operator's telephone was placed on a cradle built into the Carterfone device. A voice-operated switch in the Carterfone automatically switched on the radio transmitter when the telephone caller was speaking, and the radio was returned to a receiving condition when the person on the phone was silent.
Then in 1974 the Justice filed anti-trust against AT&T. The suit came to trial in 1981 trial. AT&T settled. On January 1st, 1984 the regional phone companies under AT&T control, the "baby bells," were divested and split into seven smaller regional bells, while AT&T kept the long lines.


When it became obvious that phone deregulation was going to happen the Cyan folks went to work. Larry Emmons always wanted to do a picture phone. Development commenced, while it would be black & white, and it only periodically sent video snapshots, it would still be groundbreaking. The project, known as Falcon, was headed up by Steve Bristow. David Stokes designed video refresh hardware. Emmons designed the echo canceling electronics, as he was purely an analog guy. Chuck Meyer did the modem.

Early prototype  

Soon Atari did embrace the project. In a 1981 Atari business plan for AtariTel, as the concept became known as,it was observed that there were 166 million telephones in the US in 1980, compared to approximately 270 million in all other countries of the world combined. In 1980 Americans use the telephone to place an average of 475 million local and over 40 million long distance calls every day.

The project was announced to the public in 1983 at the Consumer Electronics Show after nearly 2 years of working in secret. Atari had applied for trademark protection of AtariTel right before the show. The press release stated that Atari engineers had created a whole new line of consumer-electronics, which incorporates new technology and features not available in other residential telecommunications products. The release did foresee the possibility of "dramatically capable home network for voice communications, communications management, appliance control, security and environmental control." While in this age of zoom calls it hard to remember that then, this was to use plain old dial-up voice lines, which usually did not have the fidelity of AM radio. Cell phone communications was in its infancy then, and it was only analog. The digital cell network was still over a decade out.

Atari commissioned two of the world's leading industrial designers, Porsche Design and Morrison Cousins associates to develop the product's exterior design.











The AtariTel phone as marketed by Mitsubishi in 1986

The End

At its peak Cyan engineering had 30 employees. Cyan's decent started in 1983-84 when stand-alone gaming crashes and burns. Throughout the 80s Atari really had no video game success. Plus its computers were proprietary and slowly lost out to Wintel. Atari's sale to Tramiel did not include AtariTel as he did not want it. For a short while AtariTel was still intended to more forward as a separate division. Manny Gerard then decided to try and to sell AtariTel to Mitsubishi. AtariTel during that time went to Warner and tried to buy AtariTel. Warner said no to the offer as it was believed that a better Mitsubishi deal was close. In a case where one hand did not know what the other was doing, right before the deal happened, someone in HR showed up and shut the Cyan down. But Larry Emmons managed to hold enough of the operation together and Mitsubishi goes through with the deal. A number of Cyan employees went with the sale to Mitsubishi.

Mitsubishi re-branded AteriTel and sold the picture phones for a few years, at which point technology, and plain old marketing forces made the picture phone platform obsolete.

After the sale of the consumer games to Tramiel, with Warner keeping the arcade division, both pieces from there out continued to be bought and sold, merged, re-organized, etc. Much like the story of Grass Valley Company. At the beginning of 1999 the company Atari merged with Hasbo. Atari's intellectual property was sold to Hasbro interactive for only 5 million in cash. The name had already been sold to another company for home use.

Atari might have been the fastest zero to making millions to back to zero maybe ever.

Applied Design Labs

The surviving group up in Grass Valley re-constituted Cyan Engineering as Applied Design Labs (ADL). Cyan had carried out some robotics research and development work on behalf of Atari, including the Kermit mobile robot, originally intended as a stand-alone product intended to bring a beer. The company also programmed the original animatronics for Chuck E. Cheese's Pizza Time Theater pizza chain in 1977.

After Atari, as ADL is continued research work on small portable "briefcase sized" computers along with pointing devices (mice) & touch screens, before these devices became commonplace. They also continued to work with Nolan Bushnell who wanted an innovative stuffed animal, a "recording teddy bear" that could play back brief sentences spoken to it. He had just started a toy company, Axlon, and had lots of ideas but no electronics design department.

Bushnell gave ADL two tough requirements for the bear's voice:
• The voice box has to cost less than $2.50 to build
• And it had to be ready for the International Toy Fair, only two months out!"

They delivered.

ADL also hooked up with Ultra-marathon runner Jim Walker, who was frustrated with the absence of a reliable way to measure distance so he could get off the track for his workouts.
ADL combined his expertise in the field with Ron Milner's ultrasonic distance measuring technique and ADL's electronics and design experience to build an advanced pedometer. It became known as the Monitor. The concept was accepted by Nike.

ADL also developed HOT LIXX Guitar, which was voted a top boys' toy. ADL idea: combine interesting variations of music in an instrument such as a guitar for listening and performing pleasure by people without musical training.

Other Early players

There were a number companies that were started by people who had left the Grass Valley Company. The companies covered here have come and went, but many still had big effect on the area, and the television industry.

Others taking advantage of the food chain

What's Left of the Geneses

NVISION 2.0

As we pointed out at the very beginning of this book, the only physical presence left as this book is written, but it too will be gone when you read this, is what was once the NVISION building in Grass Valley. That building has been sold. But let's take a look as to how that building stayed part of the industry for as long as it did.

When we left the NVISION tale last in chapter 8 it had been sold to ADC, and in short order ADC was seriously considering just turning off the lights and walking away. On the bright side digital television transmission was now a reality, and HD television was starting to find some legs, albeit extremely wobbly, but at least it was trying to stand. But would NVISION be around to cash in?

In 2002 ADC was going broke, and they had been on a buying spree up to and including NVISION. Some of those acquisitions were not working out. The dot.com crash didn't help. On top of that, ADC's main market, telcos, was undergoing a severe downturn. They even had built a million plus square foot building in their home city of Eden Prairie, Minnesota, a southwest suburb of Minneapolis, that sat vacant due to the downturn in business. They approached Miranda; a company you saw in the last chapter that had a large effect on the Grass Valley story. But before that part of the story, a decade prior NVISION was in short, slow dance with them at this time. Miranda expressed interest. NVISION went to NAB in 2002 and was in the Miranda Booth in anticipation of a proposed sale. ADC had gotten rid of NVISION's sales force and had theirs trying to sell NVISION products. This was the exact opposite of the Wavelink problem Birney saw at GV. At GV they were TV folks trying to sell into telcos. At ADC it was telcos trying to sell into television. At this point NVISION had no market presence. ADC by itself had a very small television footprint.

But as things unfolded ADC finally figured out that Miranda didn't have any money at the time and was just trying to get a freebie. ADC then came to an improbable solution. Simply shut down the NVISION operation and walk away. Birney pointed out to ADC the warranty commitments that would be unfulfilled, this gave ADC pause. At that point, Dayton called Jim Meadlock and told him he thought ADC would take a low ball offer and he said to go for it. They made ADC a $1 million offer and they took it. Meadlock added some capital, and in June of 2002 the company under new control officially restarted.

Meadlock invested about another 2 to 2 and one-half million dollars to get marketing back up and to get back into new product development. ADC had paid the NVISION building off, so that came back technically for free. The NVISION building was originally a Blue Cross data center. The building was bought for about $750,000 and had $200,000 in improvements made. Since NVISION got a full restart, everybody got founders stock. This was a good thing because part of the deal for everyone with stock in NVISION when it was sold to ADC, with the exception of Meadlock, got ADC stock, which at this point was pretty worthless. Chuck Meyer was the president, and Birney had the title of chairman. It was NV 2.0.

At the time of the buyback Dayton said that NVISION is in the business of exporting intellectual property and importing dollars, He added that NVISION is the type of high-tech company that can make a substantial contribution to the local economy. With what was going on in the local area economically at the time it was re-assuring when he added, "We buy things from the outside, but we add a lot of value to the product before we ship it out and bring in substantially more (money) than we send out." The company had 65 employees at the restart.

Jay Kuca said that "We’re the only company here in town that can, on a volume basis, take raw PC boards, run them through our production line, and come out with a finished product on the end." That was an ability that was becoming in short supply in Grass Valley at that time.

The HD transition started in earnest in late 2002 with ESPN's and Turner's decision to launch HD cable channels. At that time ESPN also required any vendors producing their games to do it in 720p HD. By the fall of 2003, CBS and FOX joined in the fray with all the NFL rights holders producing at least their A game in HD. Once the NFL season was over, those facilities became available for spring sports and the damn burst in terms of demand for HD facilities.

The remote truck vendors faced a dilemma. Complicating this was that the networks insisted that cameras work in their chosen HD format, and not undergo conversion from the other format to their chosen format. That meant these vendors would have to have two sets of cameras, ones for ABC, ESPN, Fox, and another set for CBS, NBC, and PBS.

Technology, and a camera company that eventually saw its video operations become part of Grass Valley Company, Philips, derived a solution, that was really a compromise, but close enough to the native format requirement to be accepted. CCD Sensor density had finally got to a point where you could pack enough pixels into a 2/3 inch image real estate to sub-sample out a number of formats, vertically, from a single imager array. Simple math offered a solution; the answer was prime numbers.

A prime number is a number that can only be divided by 1 or itself. The first eight prime numbers are 2, 3, 5, 7, 11, 13, 17, and 23. If you could find a set of different row numbers that reduces to a combination of prime numbers, then no vertical interpolation is needed, simple sub-sampling would do the trick. One other requirement—you want to use the smallest prime numbers that you can, as the largest prime number used sets the overall size of the complete array.

As it turns out, the first three prime numbers are all that are needed. Pick any format's line count, and start dividing by two until you can't, then divide by three until you can't, then you'll find you need to divide once by five to end up with only prime numbers. Or start by dividing by 5 once, and then three, and then finish with 2. It doesn't matter which order you use in division. In all cases, you will arrive at the same factored result.

To hit all the required line rates, it works out that the sensor needs 4320 lines. Armed with that number of lines, you can subsample out 1080, 720, and 540, and 480 lines, the ones required. 1080, for 1080p; sample every 4 lines; 720p every 6 lines; 540, for 1080i, every 8 lines. So no conversion for the vertical line rate, but interpolation is still needed for horizontal resolution. The consensus was "close-enough." HD really took off after that.

It helped that relatively inexpensive flat panel TVs, which became as important as HD itself. They rapidly were finding homes as wives liked the look of flat panels and husbands craved big TVs, so permission to acquire was granted and the TV market grew.

But back to NVISION's first product. NVISION had a number of things that made it an early bet in the HD equipment marketplace. Besides HD being in the company's DNA, they often described themselves, and approached their designs as a RF company. From the git-go unlike other AES audio routers where the input undergoes serial to parallel conversion, then goes to processing, has the reverse parallel to serial conversion done and then out, the NV router was serial all the way through. Their solution was to buy the fastest RF parts for the output logic regeneration.

The company also always took an interest in new parts, often allowing vendors to use their facilities to test new parts. Paul Ferguson, with Analog Devices, had an Analog to Digital IC with no way to test it. He came out to NVISION as Dayton and Meyer knew ICs. A number of chip vendors came up to NVISION to test their products. So they would work with the chip vendors and get early samples and amazing pricing. Another benefit was that the company could build product faster than anyone.
Why? Did having the fab equipment help?

Sometimes being cutting edge did not always work smoothly. Early on a special IC that NVISION was counting on was discontinued. So the company had to implement an early instance of a Gate Array, with the delay in having to program the device.

Bob Adams, an AES fellow, designed sample rate conversion that would put a NV product out of business. He came out to NV for advice. What happened?

This experience helped ready the company to be the pioneer in HD that it had originally set out to be. In 98 HD router design was an art. Devices were barely working at the 1.2 GB telco rate, let alone the 1.5 GB HD speed. Printed Circuit Boards, along with connector dielectric made for very unreliable operation. HD equipment to support the HD clock rate had to be able to pass frequencies of well over 4 GHZ (1.5Gb third harmonic). Early on, until a company called Gennum produced ICs that could clean up HD serial streams by a process called re-clocking, the only other way was through a process called equalization, at a devices input and outputs. Simply put, the high frequencies were extremely emphasized at the expense of the low frequencies, knowing that for every millimeter that signal traveled it was those high frequencies that were getting clobbered. HD was a very local affair where you did not push signals any distance.

Please clarify: Bruce Regesmy, in NVISIONs test, shows a bulkhead connector blindly meets connector boards with the motherboard that greatly decreasing return loss. They called it a snap on guide, and NV got a patent for it. At the time this first router is being worked on nobody wanted active assemblies on the back plane, now it's OK.

Careful mechanical design had to be used. The mechanical packaging was paramount. The slightest change in connector alignment would render the router unreliable. The Amp Connector Company had what it called Micro Z connectors that worked at 10 GHz. But the connectors had to align within a mil of each other. Higher frequencies also equal higher heat which leads to thermal expansion, misaligning connections. Anyone care to elaborate?

NVISION used aluminum in the frame as it had a similar expansion coefficient to match the expansion of the print circuit boards as things heated up. This devotion to mechanical detail allowed the company to build HD routers before others. As mentioned, NVISION also determined that the semiconductor chip manufacturers specs weren't as good as billed. As already mentioned the chips were originally designed for 1.2 GHz telco use, and not 1.5 GB/s video use. Getting the mechanics right was so hard that as we saw in the last chapter, the NVISION router, once refined, stayed the standard for how HD routers were constructed. And as NVISION was handed off to Miranda, and then one of a number that also brought routers to the gala that Belden concocted, it faced competition in routers locally and globally, all under the same tent. A couple NVISION designed routers are still sold by the Grass Valley Company to this day. Dayton thinks that is probably the case since there is no one around today that still has mastered the art of building HD router frames. This is not a trivial problem. They would require new IO printed circuit boards, requiring hard to implement connectors, and the real hard part, a new motherboard. The motherboard is a quarter inch thick, and it has 38 layers. To make a new frame reliable requires 5/1000th of an inch tolerances across the entire frame.

Miranda Sale

Like many industries the one that this book is about, tend to take on an astrophysics bent to them. Either you accumulate enough mass to reach critical mass, with the resultant gravitational pull to have enough sway in the marketplace to sweep up sales and revenue that you wouldn't if your "star" wasn't so bright. Taking that analogy forward in evolutionary terms. You're either a large dinosaur at the top of the food chain, or you're a small mammal hiding in the underbrush, not needing much to consist on. Your advantage, if you do not get crushed, is you're nimble and much more adaptable. The problem for the giant is, back to nuclear fusion, often you construct your star, and it becomes so massive it goes supernova, and spews all your collected mass back out into the universe.

We saw it with RCA, who laid a good part of the technological foundations for the industry, become so massive that its accumulated momentum allowed Grass Valley, when it was small and nimble, along with Sony in the 80s and 90s, make the founding juggernaut irrelevant. But none-the-less, even though scaling hasn't lasted eternally yet, the desire to try again is always present. Besides Sony, other notable contestants have been Harris, and now what use to be Grass Valley, and is mostly now an international company with the name of a town that most of its current employees have never been to, and now really will have no reason to ever visit.

Grass Valley on its way to being the company it is today, came together as we saw from a number of directions. The two paths germane to the local story being the namesake company itself, and NVISION. The NVISION path finally ended up going through Miranda. Miranda decided to move up the food chain. An impetus for Miranda's move was partially due to their main investor, Société générale de financement (SGF) was a holding company owned by the Government of Quebec. One of their stated goals is to bring jobs to Canada, more exactly Quebec. In December 2010, a merger of the SGF into Investissement Québec received royal assent.

With SGF's financial backing, Miranda was serious this time about acquiring NVISION. The financial backing came with a price. Dayton felt the in dealings with Miranda it always seemed that there was no one in the company that could make a final decision, and he thought there was a puppet master. In a premonition of things to come, it was briefly discussed whether Belden should be approached as they were on a buying spree at the time. Having been bought by a similar company, ADC, once before Birney Dayton wouldn't have considered a Belden offer. Chuck Meyer wished they had shopped NVISION to others, but Meadlock wasn't a fan of that idea. He was ready to cash out. So at the beginning of 2009 NVISION sold itself for Miranda for $40m.

Miranda had sales of $37.6 million for the quarter ending Sept. 30, 2008, up 19 percent from the previous year. NVISION, reported revenues of around $34 million over 12 months to the same date. Both companies had comparative gross margins of about 60%. While Miranda was about three times the revenue of NVISION, NVISION was a powerhouse in routers, an area where Miranda needed some help. NVISION' marquee products at the time were the 8500, the 7212 routers, and a master control switcher.

Miranda's CTO, Michel Proulx, summed the acquisition up, "It's pretty clear that our product lines are very complementary. We're building all the infrastructure equipment but didn't have a router line. They have a router line." Miranda was big on interface products, and often they are needed with new routers, and visa-versa. While both sold master control, again, new master controls often require new routers. Finally Miranda was big on multi-viewers, which, again, often at the time needed new routers.

NVISION had about 100 employees in their building at the time of the sale. The headcount almost immediately started dropping as manufacturing was moved to Montreal, even though the Miranda President said router manufacturing would stay in Grass Valley at the time of the sale. Next they then moved test out of the building. Soon there was nothing left in the building except engineering. After the sale to Miranda Chuck Meyer had some title but essentially was the site manager. Part of the sale required Birney to stay a year. He recalls that Miranda never asked him a single question. Birney retired at the beginning of 2010. Chuck recalls that from an engineering perspective they managed to keep him busy.

As we saw in the last chapter, Belden then buys Miranda, and then Belden also buys GVG. This was a classic market consolidation game, buying companies up and reducing the players in the market, so you can try to control prices on your own, at least some of it. Management usually looks at a merger like it is one plus one equals three, with a longer reach, and a larger and stronger package of solutions. From the other end, accounting and almost always the people in the trenches see it as one plus one equals 1.5. The goal is always lower costs, which usually means fewer people.

The Current Big Two