Sorry for the terse posting. I should have defined a few terms. A Rydberg atom is an atom (Bucksbaum used Cs) excited to a high principal quantum number state with a high energy (negative energy close to zero). Grover's algorithm is a procedure for purifying a quantum state of all but a single component which has been marked by having its sign flipped, in effect "searching" for the sign-flipped component. The no-cloning theorem states that you can't make an identical copy of an arbitrary quantum state without destroying the original. A QC implementation is a physical system implementing a quantum computing algorithm.

Forgive my ignorance. What is a QC implementation?

How many electrons does it take to remember the entire contents of the Library of Congress? Only one, according to University of Michigan professor Philip Bucksbaum.

As usual with these journalistic accounts of cutting-edge physics, the story in EE Times is garbled and hyped. Bucksbaum's paper is available on the lanl server.

Bucksbaum is working with Rydberg atoms, and he has shown an implementation of the Grover "quantum search algorithm" (which even Grover is now calling "quantum state amplification" because it is clear that the idea of using it to search a database is absurd, given the no-cloning theorem). The fine print here is that you destroy the "library" each time you interrogate it, so you have to set it up again each time, requiring the full library's worth of classical data transfer to, in this case, googles of laser optics and so on.

The claim of inifinite data storage capacity is based on the fact that a pure Coulomb-potential atom is characterized by an infinite ladder of states below zero energy (ionization). However, the binding energy gets smaller and smaller as you populate higher and higher states in search of more and more state space to encode your "Library of Congress," and at some point the atom is going to be susceptible to any small influence that might push it over the top and ionize it. In reality, this is not a good candidate for mass data storage, although it is good experimental physics and Rydberg atoms may be a viable candidate for a QC implementation.