Presently our digital computers rely on bits, which, when charged, represent on, true, or 1. When not charged they become off, false, or 0. A register of 3 bits can represent at a given moment in time one of eight numbers (000,001,010,...,111). In the quantum state, an atom (one bit) can be in two places at once according to the laws of quantum physics, so 3 atoms (quantum bits or qubits) can represent all eight numbers at any given time. So for x number of qubits, there can be 2x numbers stored. (I will not go into the logic of all this or this paper would turn into a book!). Parallel processing can take place on the 2x input numbers, performing the same task that a classical computer would have to repeat 2x times or use 2x processors working in parallel. In other words a quantum computer offers an enormous gain in the use of computational resources such as time and memory. This becomes mind boggling when you think of what 32 qubits can accomplish.
This all sounds like another purely technological process. Classical computers can do the same computations as quantum computers, only needing more time and more memory. The catch is that they need exponentially more time and memory to match the power of a quantum computer. An exponential increase is really fast, and available time and memory run out very quickly.
Quantum computers can be programed in a qualitatively new way using new algorithms. For example, we can construct new algorithms for solving problems, some of which can turn difficult mathematical problems, such as factorization, into easy ones. The difficulty of factorization of large numbers is the basis for the security of many common methods of encryption. RSA, the most popular public key cryptosystem used to protect electronic bank accounts gets its security from the difficulty of factoring very large numbers. This was one of the first potential uses for a quantum computer.
"Experimental and theoretical research in quantum computation is accelerating world-wide. New technologies for realising quantum computers are being proposed, and new types of quantum computation with various advantages over classical computation are continually being discovered and analysed and we believe some of them will bear technological fruit. From a fundamental standpoint, however, it does not matter how useful quantum computation turns out to be, nor does it matter whether we build the first quantum computer tomorrow, next year or centuries from now. The quantum theory of computation must in any case be an integral part of the world view of anyone who seeks a fundamental understanding of the quantum theory and the processing of information." ( Center for Quantum Computation)
In 1995 there was a $100 bet made to create the impossible within 16 years, the world's first nanometer supercomputer. This resulted in the NanoComputer Dream Team, and utilizes the internet to gather talent from every scientific field and from all over the world, amateur and professional.
The possibilities for making a nanotech quantum computer are many, including such exotic creations as quivering nanotubes, superconducting nanocircuits and quantum dots.
"Nanoscale devices are the best case to observe quantum mechanical phenomena in the compromise between something small enough to be quantum mechanical, but still large enough to be controllable and accessible," said physicist Franco Nori of the University of Michigan at Ann Arbor and the Frontier Research System of RIKEN near Tokyo.
Conventional computers work by symbolizing data as a series of ones and zeros - binary digits known as bits. The resulting binary code is conveyed via transistors - switches that can be flicked either on or off to represent one or zero.
Quantum computers, however, take advantage of the strange phenomenon that physicists call "superposition," where infinitesimal objects such as individual electrons or atoms can exist in two or more places at once, or spin in opposite directions at the same time.
This means computers built with superposition processors could employ quantum bits - called qubits - that exist in both on and off states simultaneously.
Quantum computers therefore can calculate every possible on-off combination at the same time, making them dramatically faster than conventional data processors when it comes to solving certain problems involving probabilities, such as code-breaking.
Quantum computing research is growing rapidly at military, intelligence and university research labs worldwide, as well as at those of industrial giants such as AT&T, IBM, Hewlett-Packard, Lucent and Microsoft.
To run mind-boggling calculations, scientists will need to scale up quantum computers from the handful of qubits most now possess to hundreds. This will be difficult, because superposition is an extremely delicate state of matter that can be disrupted by the slightest disturbance.
So far, scientists at best have managed to link up, or entangle, only a few qubits to perform simple logic operations, Nori said.
The first experiments that created qubits used particles such as chloroform molecules whose components were pushed into superposition with magnetic fields and radio waves. Among the problems with such devices was they did not scale up qubits readily.
That is where nanotechnology comes in.
"As a result of the existing semiconductor industry, a great deal of expertise is available for micro or nanofabrication," said physicist Albert Chang at Duke University in Durham, N.C.
"This knowledge base could greatly short-circuit the design and implementation of multi-qubit circuitry."
Among the most promising candidates for quantum computers are quantum dots - semiconductor crystals only nanometers, or billionth of a meter, long. Scientists can cram electric charges into quantum dots so they behave like puddles of electrons.
The key to using quantum dots in quantum computing is their property known as spin. Electrons spin just as Earth spins on its poles. When two electrons occupy the same space, they must possess opposite spins - one electron spinning "up" and the other "down," Chang said.
Electrons can even be packed into quantum dots so each dot has a net spin of up or down.
Chang and colleagues created qubits from quantum dots by placing a pair of dots carrying the same net spin value near each other. They connected the dots with tiny wires and directed how much electric charge the dots could transfer among one another.
By controlling the charge transfer, the team converted both dots to qubits, spinning both up and down simultaneously.
"In my view, in the longer run - say on a five-to-10-year horizon, quantum dots have a good chance to emerge as one of the best systems," Chang said.
In addition to Chang's group, other notable researchers working on quantum-dot qubits include Charles Marcus at Harvard University, Leo Kouwenhoven's group at the Delft University of Technology in the Netherlands, Seigo Tarucha at the University of Tokyo and Jorg Kotthaus at LMU Munich.
Similar to quantum-dot computers are Kane quantum computers, named after physicist Bruce Kane who suggested the idea in 1998 when he was at the University of New South Wales in Sydney.
In a Kane quantum computer, phosphorus atoms under a layer of silicon 25 nanometers or so deep behave as qubits. The device uses phosphorus because the atoms can remain in superposition for a long time.
The Kane quantum computer represents the primary quantum-computing effort in Australia. Because it also depends on silicon, the hope is techniques long refined in the semiconductor industry will help to manufacture these computers and scale them up to large qubit numbers.
Still another major contender uses superconducting nanocircuits, which government and university labs worldwide are researching.
At the nanolevel, electronic circuits begin to exhibit quantum behavior. In superconductors, electrical current flows with no resistance, which means electronic signals can travel without energy loss, helping to preserve superposition.
Scientists have worked on superconducting devices for roughly 40 years, and in many ways the superconducting approach for quantum computers is very advanced compared to others, explained physicist Andrew Cleland of the University ofCalifornia,Santa Barbara.
Still, the circuits are currently prone to having the superpositions break down, Chang said. A key question that remains to be answered is whether error-correction schemes can overcome this problem to make superconducting nanocircuit-based quantum computation "a practical and useful reality," he added.
A more robust quantum computer might even prove mechanical in nature, Nori said. He and colleagues recently proposed using carbon nanotubes or silicon nanorods as mechanical qubits.
"Imagine a ruler and squeeze it along its length," Nori said. A normal ruler would bend either left or right, but if shrunk to nanoscale dimensions such a ruler would take on a superposition of buckling left and right at the same time.
The advantage of a mechanical qubit is it potentially could remain in superposition longer than other kinds of qubits. Moreover, mechanical qubits could be manufactured via simple carbon-nanotube growth techniques researched feverishly the world over, said Nori's collaborator, physicist Alik Kasumov of the University Paris-Sud.
"Isn't the basic idea the coolest thing?" asked Keith Schwab, senior physicist for the National Security Agency's lab at the University of Maryland in College Park.
Kasumov, Nori and colleagues plan experiments on buckling nanotubes this year, and mechanical qubits could appear within three years if they can produce superposition in the nanotubes, as hoped.
No comments:
Post a Comment