What is a Quantum Computer?

 by

 Sam Stephens

It is obvious to anyone on a college campus full of iPod-carrying undergraduates that computers are getting smaller.  An iPod is a small hard drive that would never have been created had scientists and engineers not tried to create smaller and smaller devices used to harness energy and represent information.

Dating back to transistors – the devices in your electronic equipment that control the flow of electricity – computers have used materials so small that they cannot be described by the laws of classical physics.  A branch of theoretical physics called quantum mechanics is necessary to explain how the materials behave.

Quantum mechanics was devised to explain happenings at the atomic level, such as the ability of electrons to orbit an atom’s nucleus without collapsing into it. Some particles are so small we cannot measure them precisely without altering them, and quantum mechanics describes these particles with something called a wave function, which is sort of a record of all the places it is likely to exist.   

What if the many possibilities for a particle’s location could be encoded as information, though?  Could there be a way to create a computer that works on multiple problems at the same time?

Theoretically, yes, and the concept is called a quantum computer.  To understand how quantum computers work, it helps to have an understanding of how classical computers work.  The basic unit of our current computers is called a bit, a magnetization or charge represented by a zero or a one.  Information that is put into a computer is coded in series of zeros and ones, and engineers construct systems that manipulate electric charges that correspond to the zeros and ones.  A one means a charge would be present and a zero means it would be absent, for example.

Engineers also construct logic gates, structures that turn the charges on and off according to the laws of logic.  For example, a “NOT” gate takes in a bit as a one and spits it out as a zero, and vice versa.

Quantum computers would store information in quantum bits or "qubits", rather than bits.  Qubits can be represented as zeros and ones, but they can also be represented by a combination of the two called a superposition.  The reason a qubit can be in two states at once is because it is a quantum system to which principles of quantum mechanics apply–so its location is described by a wave function, which describes the likelihood of it existing in different locations.

Jose Villas-Boas, a post-doctoral fellow at Ohio University, collaborated with professors Sergio Ulloa and Alexander Govorov on a paper published in February 2005 describing how to more effectively manipulate quantum dots, which theoretically can be used in quantum computing.  Quantum dots, also known as artificial atoms, are crystals that are approximately the diameter of about 10 to 50 atoms and can contain 100 to 1,000 electrons.

The electrons are important because they can be sent through transistors, several at a time, after the dot is bombarded with light from a source like a laser.  The fact that more than one electron could be processed by the transistor at once is the main innovation that quantum computing offers, Villas-Boas said. 

“In a classical computer, you have to insert one question at a time,” he said. “The big difference is that in a quantum computer instead of zero and one you can have any combination of zeros and ones.”

Let’s think about why we would ever want to harness the powers of superposition.  Although people who are instant messaging someone in Brazil while listening to a radio station from Kenya on their personal computer might think otherwise, there are some things our computers can’t do.  Factoring extremely large numbers, for example.  Sure, a program can be written for the task, but there comes a point when there is simply not enough time to run that program for every possible combination of zeros and ones to find the correct answer.

Quantum computers would best be utilized in cryptography and as a simulators that would create models of quantum physical systems, said Kai-Felix Braun, a research assistant professor of physics at Ohio University.

“I would not say it has advantages for everyday applications,” he said.

It is difficult to envision a quantum computer with any more than a few qubits for its foundation, so let’s imagine a two-qubit quantum computer.  There would be two small particles, such as two electrons or photons. A two-bit classical computer can exist in four states: 00, 01, 10, and 11, whereas the two-qubit quantum computer would exist in those four states with various probabilities.

To understand what the probability represents, it helps to understand what happens when the state of the quantum computer is measured, a phenomenon called decoherence.  The name decoherence comes from the idea that a quantum system collapses and becomes governed by classical physics laws if it has a certain interaction with its environment.  It is extremely difficult to measure the state of a quantum computer without making such a disruption.

“One problem with these qubits is that they hold their state only a certain time,” Braun said. “Also, if you want to bring the information from one qubit to another it might get lost.”

This problem could be solved by harnessing a property called electron spin, Villas-Boas said. Electrons spin one of two ways, analogous to the clockwise or counter-clockwise spinning of a top but coded by scientists as “up” or “down,” which could be analogous to “0" and “1" when storing information.

Because the spin of an electron is relatively insensitive to its outside surroundings, it takes a long time for those outside surroundings to lock it into decoherence, Villas-Boas said.

 “It’s hard to talk to it,” he said. “It does not see the world around it.”

This is advantageous to the development of a quantum computer because it allows the computer more time to make calculations before decoherence happens and the state of the computer becomes fixed in space and time.  The disadvantage to using electron spins is that manipulating the spins is difficult.

“How can you operate it if you can’t talk to it?” Villas-Boas said.

Think back to the two-qubit quantum computer.  If you send two qubits through quantum logic gates, the gates are actually operating on four states at once because they take advantage of the quantum mechanical properties of the qubits; think of them as 00, 01, 10 and 11.  When you finally “talk to” the system and make a measurement, the computer collapses into one of the four states–the output of the logic gate and thus the “answer” to the “question” the logic gate is asking. 

Only when we think of larger quantum computers does this concept begin to seem useful, though.  It’s a theoretical concept that has many unsolved practical problems, but you can begin to see the benefits associated with quantum computing.  A string of only 25 qubits, for example, can exist in 2²⁵ different states, which is more than 33 million states!  Instead of the impractical drudgery of sending each possible state one by one through a logic gate to find out which one is the correct “answer,” the whole 25-qubit quantum string is sent through and operated on by the quantum computer at once, and the quantum computer logic gate produces the answer after evaluating all 2²⁵ states simultaneously! This would be a very powerful tool if it could be built.

The road to useful quantum computers is littered with obstacles, some of which we discussed.  But the idea of taking advantage of quantum properties to send and process information more quickly is intriguing, to say the least.