Even with successful coin-ops using new technology, Atari was still finding the competition for arcade floor space suffocating. With a $250,000 cost to develop a game, and about 10% chance it would be successful, Atari had to become serious about their second front: the home.
However, that would not be easy. After being successful in 1975 and early 1976 with home Pong, sold through Sears, Atari created more versions of home Pong to market both themselves and through Sears including C-140 Super Pong, C-160 Pong Doubles, C-180 Super Pong Ten. However, by that time, Atari was not the only company selling a chip-based Pong unit. At least 75 companies announced their intentions to get into the business in early 1976. Atari might have made the first home Pong, but just like its experience with their coin-op Pong, dozens of imitators were in pursuit of the same dollars. Atari tried to combat the imitators by generating a sense of loyalty in consumers to buy "A Real Atari", but compared to some of the newer models, Atari’s home units looked primitive just 6 months after they went on sale.
With the focus on home video games and coin-ops, Atari did make one move in 1976 that, in hind-sight, could have been their biggest mistake. Steve Jobs, who had left Atari and was working for HP at the time, brought a piece of equipment to Al Alcorn that he and Steve Wozniak had been working on in their garage. It was a computer based on the same MOS 6502 processor that Atari had started using in their coin-op games. Alcorn thought it was a “neat engineering project” but did not think it was right for Atari.
“we said ‘no thank you’…but I liked him, I thought he was a nice guy, so I introduced him to venture capitalists’”lxxx - Al Alcorn
Atari did not have the resources enter a third, unproven market. They let the opportunity go. The machine became the Apple I computer, and rest is history for Jobs and Wozniak. Atari learned to regret this mistake as they tried to enter the home computer market a few years later.
Bushnell was convinced that Atari needed to outthink their competitors -- and could. The 1975 hit coin-ops Tank! and Jet Fighter were suggested as dedicated follow-ups to the Pong units, but Atari was tired of designing and selling dedicated hardware that cost $100,000 to developlxxxi with only two to three months of shelf-life before it became outdated. The company needed to design a platform that could sustain a life of two or three years, and at the same time support incremental game sales to an already established user-base. Months before the Fairchild Channel F appeared on store shelves, Atari was well on its way to creating a similar, yet much more flexible system.
The R&D team at the Grass Valley Think Tank started working on ideas for a programmable unit that could use interchangeable games as early as mid-1975. The problem was, most suitable microprocessors, like the Motorola 6800, cost $100 each. This was too much for a consumer product. In September 1975, Steve Mayer and Ron Milner met Chuck Peddle (who had recently left Motorola) at Wescon and made a deal to buy his microprocessors (MOS 6502) for $8 apiece. The 6502 met the minimum specs required for the reprogrammable system they were planning to create, and could also be used as the basis for microprocessor based coin-ops.
In December 1975 Atari R&D at Grass Valley hired Joe Decuir, and one of his first projects was to help debug a piece of hardware that would become the Atari VCS console.
“Steve Mayer and Ron Milner conceived of the VCS, and designed the first prototype of its ancestor.”lxxxii - Joe Decuir
Soon, the VCS project attained the name “Stella”, named after Decuir’s bicycle. The first programmable system prototype used controllers from Kee’s Tank coin-op, the custom Stella chip, a 6502 hobbyist board, and a 5 volt power supplylxxxiii.
A second prototype was developed in Los Gatos in March 1976lxxxiv, as Joe Decuir worked as an apprentice to Jay Miner, a legendary Silicon Valley hardware designer, and the only person Al Acorn knew of who could pull off the project. They set out to create a machine whose inner workings were accessible to the programmer, and could be exploited by those who got to know the hardware well.