Scientists at MIT
and Los Alamos National Labs recently announced that they've developed the first
quantum computer able to simulate a quantum system. The prototype can only count
to 4, but its creation is a milestone on the road to real quantum computers.
After languishing
for more than a decade, quantum research gained momentum in 1994 when Peter
Shor, a scientist at AT&T Labs, showed that a quantum computer could be
programmed to break an important kind of code exponentially faster than existing
computers. The floodgates of research interest opened all over the world,
leading to a recent breakthrough.
The Nobel
Prize-winning physicist Richard Feynman suggested in 1982 the possibility of a
new kind of computer that would exploit the quantum properties of matter. Such a
computer would represent data as quantum bits (qubits) instead of
ordinary bits. Qubits aren't limited to being either 0 or 1. Rather, they can be
in a mixed state of on and off. The result: built-in massive
parallel computing.
Quantum computers
are uniquely powerful because qubits can interact with each other, giving rise
to exponential increases in power. In conventional computing, the power of a
processor increases additively with each bit, but in quantum computing, an added
qubit multiplies the potential power of the processor, doubling or even
quadrupling its capability.
A Simple
Experiment
The group of researchers, including Raymond LaFlamme of Los Alamos, and David
Cory and Ching-Hua Tseng of MIT, have developed a general framework for quantum
simulation that could be adapted to any quantum computer. Tseng explains that
the experiment they did was "very simple...a first-year quantum mechanics
student could do it on paper. But this is probably the first reachable
application of information processing on a quantum system."
Cory says that
we're at the very beginning of the era of quantum transducers but
cautions that "we're still in the early stages of integrating information
theory into quantum mechanics." Nonetheless, he suggests that there may
even be a quantum improvement to Moore's Law, which refers to the doubling of
computer power every eighteen months. "In the last two years we've gone
from two qubits to six qubits. This is a sixteen-fold increase in computer
power. We may have 10 qubits in 2001, another sixteen-fold increase."
Even adding two
qubits a year, quantum computers are years away from rivaling existing
supercomputers. A 40 qubit machine would be more powerful than even the most
powerful existing computers. A comparison of quantum computers to other future
computing paradigms such as molecular electronics is even more revealing: Cory
explains that "...all you need are 86 qubits and you have a quantum
computer that is more powerful than....a classical computer [built from] every
molecule of water on the earth."
There is an irony
in all this. Johnny von Neuman developed the theory of classical computing,
built into the logic circuits of every (non-quantum) computer. He also
formulated the mathematical foundations of quantum theory. In the 1930s, he and
Garrett Birkhoff developed what they called 'quantum logic' as a framework for
understanding the deeply mysterious laws of nature at the submicroscopic level.
This logic is embodied in today's quantum computer prototypes, and it will be
the basis of the quantum computers of the future. But, as Peter Shor, recently
named a MacArthur Fellow, explains, "von Neuman never made the connection
between quantum logic and computing. It took 50 years and Richard Feynman to
bridge the gap."
