I assume John Markoff knows that matter is made of atoms, and thus that transistors cannot be made "infinitely" smaller for that reason alone. I assume that poor editing is what makes it appear that Markoff is saying, "keep the insulating regions large enough, and transistors can be made infinitely small."
But it is funny...
Posted by DeLong at December 15, 2002 06:33 AM | Trackback
I.B.M. Plans a Tiny Transistor By JOHN MARKOFF | SAN FRANCISCO, Dec. 6
AN FRANCISCO, Dec. 6 — A team of I.B.M. researchers plans to report on what would be the world's smallest transistor when they deliver a research paper this week at the annual International Electronic Device Meeting opening here on Monday.
The researchers said the advance would help ensure that the semiconductor industry could fulfill its performance projections through at least 2016.
The transistor, which was designed by a team of I.B.M. researchers led by Meikei Ieong, is just nine nanometers in length — or at least one-tenth the size of the most advanced transistors currently in production. A nanometer is one billionth of a meter; the average human hair is more than 3,000 times as wide as a nanometer.
The tiny transistors would allow for high-capacity memory and faster digital logic abilities. But they also create challenges because as transistors are packed ever more densely on the surface of silicon wafers, heat becomes more of a problem.
Each year a consortium of international semiconductor groups releases an International Technology Roadmap charting the progress the industry has made toward scaling down the basic components of microelectronic devices. The 2001 road map indicated that to continue at the present rate, the transistor would need to be smaller than nine nanometers by 2016.
Dr. Ieong said the ability to make such ultrasmall transistors was essential for studying the odd physical world at the subatomic level that the semiconductor industry was now approaching. There is no firm evidence that chips can be made infinitely smaller because of instabilities, like the leakage of electrons across boundaries within the device that are supposed to serve as insulators.
Sometimes "indefinitely" makes sense, and sloppiness somewhere puts "infinitely" in instead. You can keep getting more toothpaste out of a tube indefinitely.
Posted by: P.M.Lawrence on December 15, 2002 03:06 PMAnd to be fair matter is made out of atoms, which are made out of quarks, and at sizes smaller than the Plank length, well, we really do not know, do we?
Of course, we would have to assume that electrons would no longer be the medium. On another note, quantum transistors( would they be called 'transistors'?) might perform many more, but different types of, calculations. Does an increase in what is calculable count as making the transistor smaller?
Of course, though, Markoff's thought was on the wavelength that you are correctly criticising him for-
Maybe "infinite" is in the process of losing its "hard" meaning. The same thing happened to the word "ultimate," originally connoting "final." (A sportswriter at one Super Bowl asked Dallas Cowboys' running back Duane Thomas how it felt to be playing in the "ultimate" game. Thomas asked him back "If this is the ultimate game, how come we're playing again next year?")
Posted by: Jeffrey Kramer on December 16, 2002 12:21 AM