IBM Corp, which spent US$5.2 billion on research and development last year, for the first time has cracked a simple digital code using a computer built on individual atoms. The advance, in a field called quantum computing, was reported Wednesday in the journal Nature.
The largest computer maker said the research proves certain techniques that may be useful as scientists develop quantum computers, though it isn't a breakthrough by itself, the technology allows atomic-scale computers to be used commercially. Today's fastest supercomputers would take years to factor the kinds of large numbers used in modern cryptography.
In 1994, AT&T Corp researcher Peter Shor theorized that a quantum computer, because its processors are so tiny, could operate at the speed needed to factor such numbers and break codes quickly. In IBM's research, a quantum computer factored the simple number 15 -- a step toward breaking down the factors of much larger numbers used in encryption. In solving Shor's cryptography problem, IBM used pulses of radio-frequency energy to manipulate atoms so their nuclei would spin in a controlled manner. Nuclear magnetic resonance, the same technique used as a hospital diagnostic tool, helps measure the spin.
IBM synchronized the spin of seven atomic nuclei in billions of molecules in a test tube to demonstrate Shor's Algorithm by calculating the factors of the number 15. The factors are 3 and 5.
Quantum computing, though not designed for such tasks as word processing, is well suited to massive calculations.
"It's definitely a step forward," Shor said in an interview.
"It happened quite a bit sooner than I thought. They were fairly clever."
Spinning a nucleus upward represents the binary 1, and a downward spin the binary 0, creating the basics of digital computing. For the experiment, IBM custom-designed molecules containing five fluorine and two carbon atoms.
Because a quantum computer operates on an infinitesimal scale, its speed would be sufficient to factor large numbers used in today's cryptography. Though conventional cryptography would be rendered obsolete by a quantum computer because it could easily break the code, such a device could also create higher-level encryption that would be impossible to crack, said Nabil Amer, manager of IBM Research's physics group.
The task now for quantum computer researchers is to find a new way to manipulate and measure the atoms, because nuclear magnetic resonance probably won't be useful beyond the spinning of 10 atoms, Amer said.
IBM researchers and other scientists have started experimenting with techniques to control the spin of 1,000 atoms, for example. The challenge is to decide whether to use superconductors, or materials that easily conduct electrons, or semiconductors, which don't, Amer said.
RSS