The idea of sending information from one point to another without expending any energy to speak of seems reminiscent of perpetual motion machines and other engineering fantasies. Ever since 1948, when engineer Claude Shannon established the minimum amount of energy it takes to send a unit of information, engineers have assumed the matter was closed.
Not so, says ibm physicist Rolf Landauer. The conventional view of a communication link as a pipeline, in which bits--the ones and zeros of computer data--are loaded in at one end and unloaded at the other, is inherently wasteful, he says. Wouldn’t it be more efficient to recycle those bits instead? If I use some crazy machinery, I can send a bit from here to there with as little energy as I want, Landauer says.
Instead of a pipeline, Landauer proposed this past year, researchers should start thinking about designing a communications ski lift. Essentially, it would be a continuous stream of particles that loop from the transmitter to the receiver and back. Say you were using ammonia molecules, composed of three hydrogen atoms with a nitrogen atom sticking out to the side like an arrow. At first all the nitrogens would point to the left, corresponding to a zero, but the transmitter would flip some of the molecules so they would point to the right, corresponding to a one, in such a way as to encode the message. At the receiving end, the molecules in the ski lift would interact with another string of molecules, swapping orientations and thereby delivering the message before returning to the transmitter.
So why doesn’t this require any energy, or only an arbitrarily small amount? For one thing, Landauer assumes the machinery controlling his ski lift is frictionless. For another, the ammonia molecules return to the sender, so the energy contained in their mass--the energy given by the formula e=mc2--isn’t lost either. And finally, the transfer of information at the receiving end requires no energy, because it would occur through tunneling--the quantum mechanical phenomenon by which a particle can switch states spontaneously, tunneling through an energy barrier rather than having to jump over it. In Landauer’s scheme, the barrier surrounding an ammonia molecule in the receiver would be lowered just enough and at the right instant to encourage the molecule to tunnel into the same state-- left- or right-pointing--as a molecule arriving on the ski lift.
In practice, of course, Landauer doesn’t know how to control tunneling that finely, and he doesn’t know how to build a frictionless ammonia ski lift. And with conventional communications pipelines growing ever cheaper and more capacious, there is no great market just now for a more energy-efficient approach. As the demand for the number of bits we send around the world keeps increasing, there will be a time when people would like to save this energy, says Landauer. But we haven’t gotten there yet. He sees his idea less as a blueprint than as a suggested line of research--he calls it a want ad, which is different from a fantasy.