Slashdot Mirror
IBM Creates 1st Single Molecule Computer Circuit
Llowfyr writes "Yahoo has reports that IBM researchers have created the first ever single molecule computer circuits which may someday lead to a new class of smaller and faster computers that consume less power than today's machines. The IBM team made a `` voltage inverter '' -- one of the three fundamental logic circuits that are the basis for all of today's computers -- from a carbon nanotube, a tube-shaped molecule of carbon atoms that is 100,000 times thinner than a human hair. IBM scientists will present the achievement today at the 222nd National Meeting of the American Chemical Society being held in Chicago and it appears in the web edition of the ACS' journal Nano Letters."
Gain is a common figure of merit of transistor-based amplifier circuits. The gain of a voltage amplifier is defined as the ratio of the size of the output signal to the size of the input signal. An amplifier that could take a 0.5V amplitude sine wave as its input and produce a 5V amplitude sine wave as its output has a gain of 10. You don't get something for nothing, of course - the amplifier has to be connected to an external power source.
A transistor is a three-terminal device. In a typical computer chip, these three terminals are called the source, the drain, and the gate. For a given voltage between the source and drain, the current that flows into the drain is strongly dependent on the voltage applied to the gate. That's what allows transistors to be used as switches: you can make a transistor that won't let current flow from source to drain unless the gate voltage is turned up past some value.
Achieving actual gain in a single-molecule device is important. Without gain greater than one, it's not possible to efficiently chain large numbers of transistors together to manipulate signals. A strong input would get degraded with each stage of transistor manipulation, eventually falling to a level too small to drive subsequent transistors.
There are *many* problems with the idea of using individual molecules to replace Si devices. Achieving a gain > 1 is a necessary but by no means sufficient step for eventual molecule-based computers. As a physicist, I think it's important to recognize real achievements in this field, but not to buy into the hype unquestioningly.
I found the full paper here (that's http://pubs.acs.org/hotartcl/nalefd/nl015606f_rev. html for you paranoid types).
I was just thinking - they say their NOR gate is the size of approx. 1/100,000th the width of a human hair. Well, today's 1.4 GHz chips contain ~22 million transistors. That would make it 220 human hairs wide. That's a lot of power in a small space. I can't wait till the day I can crack RC5 on my cell phone.
--
#nohup cat