ABSTRACT
In the early 1990s an integrated circuit first made in 1969 and thus ante dating by two years the chip typically seen as the first microprocessor (Intel’s 4004), became a microprocessor for the first time. The stimulus for this piece of industrial alchemy was a patent fight. A microprocessor patent had been issued to Texas Instruments, and companies faced with patent infringement lawsuits were looking for prior art with which to challenge it. 2 This old integrated circuit, but new microprocessor, was the AL1, designed by Lee Boysel and used in computers built by his start-up, Four-Phase Systems, established in 1968. In its 1990s reincarnation a demonstration system was built showing that the AL1 could have oper ated according to the classic microprocessor model, with ROM (Read O nly Memory), RAM (Random Access Memory), and I/O (Input/ O utput) forming a basic computer. The operative words here are could have, for it was never used in that configuration during its normal life time. Instead it was used as one-third o f a 24-bit CPU (Central Processing Unit) for a series of computers built by Four-Phase. 3
Examining the AL1 through the lenses o f the history o f technology and business history puts Intel’s microprocessor work into a different per spective. The differences between Four-Phase’s and Intel’s work were industrially constructed; they owed much to the different industries each saw itself in. 4 While putting a substantial part of a central processing unit on a chip was not a discrete invention for Four-Phase or the computer industry, it was in the semiconductor industry. Although the AL1 was in many ways technically superior to Intel’s first generation microprocessors, its location in the computer industry led to a different, and ultimately truncated development trajectory. Flexibility was the hallmark o f Intel’s microprocessor, with Intel and its customers finding countless applica tions for it, while the industrially constructed rigidity o f the AL1 limited its applications.
