DRAM Inventor Wins Kyoto Prize

This year’s Kyoto Prizes included an Advanced Technology Prize for the father of DRAM, IBM’s Dr. Robert Dennard.

The Kyoto Prize, one of the world’s most prestigious accolades, is an international award bestowed once a year by The Inamori Foundation to honor those who have contributed significantly to the scientific, cultural and spiritual betterment of humankind.  Some say it is similar to the Nobel Prize, and seven Kyoto Prize laureates have gone on to win the Nobel Prize.

In addition to the kudos of receiving this honor, Denning was also given a 20-karat gold medal, a certificate, and a cash prize of 50 million yen (about $600,000).

Dennard invented the concept of a one-FET one-capacitor DRAM in 1967 after marveling at a workmate’s very simple core memory design.  At the time Dennard was designing six-transistor MOS SRAMs, and was looking for a way to simplify the design; his goal was to reduce costs enough that semiconductor memories could compete against core memories which, at that time, were the main storage medium for computer main memories.

His concept was to use a capacitor for storage, and control it with a single MOS transistor.

Although he was issued a patent for the concept in 1968 (US Patent 3,387,286: “Field Effect Transistor Memory”), commercial DRAM production did not commence until 1973.

This was not the first commercial DRAM.  A 3-transistor 1Kb part was marketed starting in 1970.

Consider this: In the 40 years since the introduction of the first commercial DRAMs a total of approximately 150 pentillion DRAM bits have been shipped.  (A pentillion is 18 zeros, with six commas, tacked onto that number.)  How many patents have been put into that kind of volume production?  This is a testament to the power of semiconductor scaling, which allows today’s chips to contain more transistors than engineers could even dream of a few decades ago.  Today, a single 4GB (gigabyte) DRAM module contains more DRAM bits than were shipped in the entire first year of DRAM production.

The prize was not given to Dennard solely for his invention of the single-transistor DRAM – this researcher also produced seminal research projecting the scaling trend that semiconductors eventually followed.  Dennard and his team concluded that transistors could shrink in all three dimensions leading to improvements not only in cost, but also in speed and power.  This team developed guidelines, now called “Scaling Theory” that show how to reduce device dimensions to achieve matching improvements in speed and power.  This is called “Dennard Scaling” in his honor.

Most importantly, he found that the power consumed by MOS transistors scaled with their size, so that increasing the number of transistors on a chip by simple scaling would not cause the chip’s power dissipation to increase.  Every new generation chip would have twice as many transistors at the same cost and the same power consumption.

Although Dennard is now 81, he continues his work at IBM, where he has been for his entire career.  He is an IBM Fellow at the Thomas J. Watson Research Center in New York.  His other awards include the U.S. National Medal of Technology, and the Charles Stark Draper Prize from the U.S. National Academy of Engineering.