The first Microprocessor was born 50 years ago (the Intel 4004). I suspect it would not have been possible to conceive of the possibility of the Microprocessor fifty years earlier back in 1971. It would take another 25 years before the first computer would be developed in 1945. Thinking about this made me wonder if we will still have microprocessors in 50 years, and if not, what will we have as a substitute. I have a few ideas but that will be the subject of a future post.
The Intel Alumni Organization just held a webinar on the 50th anniversary of the 4004. While a bit long, it is worth hearing about the birth of a technology that has changed the world, the Microprocessor, from three key people who made it possible. You can view it here. Please note that the opening piano music is me improvising. You can hear the full piece here.
I write about how Intel came to create the first Microprocessor in my book, The Flight of a Wild Duck. It is a fascinating story, and like many success stories, it almost did not happen. Bob Noyce and Gordon Moore founded Intel in the summer of 1968. Its objective was to develop semiconductor memories that could replace the memory technology used by computer manufacturers and others called Magnetic Core Memory. The company was started with just a $3 million investment. Bob and Gordon knew it would take time for the memory business to develop. First, they had to create some critical technologies needed; then, they had to build memory products based on these technologies. Finally, they had to convince customers to take the risk of using new products based on new technologies from a new company to replace one of the most critical components in their products. So to pay the bills, Intel decided to do custom design work for companies. The first major customer was a Japanese calculator company called Busicom who approached Intel in April of 1969 and asked the company to develop twelve specialized chips for a desktop calculator. Intel accepted the request. Intel had recently hired an engineer from Stanford University who had a computer background, Ted Hoof.
Bob Noyes asked Hoof to help manage the relationship with Busicom. In preparing for that assignment, Hoff reviewed the twelve chips requirements. He soon realized that Intel did not have the engineering talent to pull this project and told Noyce that. Noyce asked Ted if he had any ideas about simplifying the design. That is when Hoff had a breakthrough. He realized that it would be possible to develop a microprocessor with some support chips that could do all the functions needed for the calculator using just software. Others had realized the potential of a microprocessor earlier but thought the technology required was still several years ahead. While Intel gets the credit for the first Microprocessor, it was inevitable that such a product would be created. Texas Instruments developed a microprocessor, the TMS 1802NC, around the same time.
Busicom had an exclusive license to the 4004 and supporting chips. Fortunately for Intel, Busicom was having some financial problems. Bob Noyce met with their management and was able to get them to release the exclusivity by offering lower prices.
Intel needed to convince the Busicom management to abandon their design and go with the microprocessor concept. Busicom eventually agreed, which is pretty astonishing to me, given how conservative Japanese companies were. Too bad they did not ask Intel to pay them a percentage of the sales of future microprocessors.