One of the most important inventions of modern times had its origins in a quality control issue. In fall 1969, Intel was having reliability problems with its 1101 static read-only memory (SRAM), the first mass-produced chip to use metal-oxide semiconductor technology. Dov Frohman, who had recently come to Intel from Fairchild Semiconductors, was asked to investigate the problem. His findings would lead him to invent erasable programmable read-only memory (EPROM).
Frohman discovered that in certain circumstances the silicon dioxide insulators in the 1101 were absorbing electrons from the metal components and holding an electrical charge, which was interfering with the device’s function. He set to work on preventing the issue, but he also came to believe the phenomenon he discovered could be useful. If he could induce and control the silicon dioxide’s ability to hold an electrical charge, he could create a chip that could be programmed electrically, but that would also retain its information without a continuous power supply. Moreover, the chip could be erased through exposure to a force that neutralized the charge — Frohman settled on ultraviolet light — making it erasable as well as nonvolatile.
Frohman debuted his prototype EPROM in February 1971 at the Solid-State Circuits Conference in Philadelphia. Gordon Moore recalled the demonstration:
Dov projected a film that displayed the bit pattern in the EPROM memory cells. As the cells were exposed to ultraviolet light, the bits dropped out one-by-one until all that was left was the familiar Intel logo. … The bits fell, and when the final one disappeared, the entire audience broke into applause. Dov’s paper was voted the best at the conference.
EPROM resolved an issue that had been plaguing developers for years — the design time for new chip prototypes was too long. EPROM’s combination of programmability and stability reduced design time for new chip prototypes from days or weeks to hours. However, the full potential of the device was not yet clear in 1971. According to Frohman, it was initially seen as a niche device for developers: “The thought was that you would want a read-only memory (ROM) as a full production device, but you could use a re-programmable memory (e.g. the EPROM) in order to see if your design was right.”
That perception changed because of another Intel breakthrough introduced later in 1971: the microprocessor. EPROM allowed microprocessors to be easily programmed and reprogrammed, which was crucial for their use. Intel’s EPROM business would grow in conjunction with its microprocessor business until the 1980s. And when EPROM eventually gave way to other forms of memory, those new forms would build on Dov Frohman’s breakthrough.