Most people will be familiar with Moore’s Law which states that the number of transistors it’s possible to get on a microprocessor doubles every 18 months. If this holds true it means that some time in the 2020s we’ll be measuring these circuits on an atomic scale.
You might think that that’s where everything comes to a juddering halt. But the next step from this is the creation of quantum computers which use the properties of atoms and molecules to perform processing and memory tasks.
If this all sounds a bit sci-fi, it’s because practical quantum computers are still some way in the future. However, scientists have already succeeded in building basic quantum computers that can perform certain calculations. And when practical quantum computing does arrive it has the potential to bring about a change as great as that delivered by the microchip.
How it works
Like any other computer a quantum machine works on Turing Theory, using a string of zeroes and ones to give the machine its instructions. However – and this is where it begins to get a bit freaky – a quantum computer isn’t limited to just two states. Quantum bits (qubits) can be both zero and one, and all stages in between, at the same time – something known as parallelism. Qubits represent atoms, ions, electrons or photons and their control devices, all working together to act as computer memory and a processor.
What this parallelism means in practice is that while a conventional Turing Machine processor can only perform one task at a time, a quantum machine can perform many tasks at once. They this have the potential to be hugely more powerful than even the largest current supercomputers.
To put that into numbers, today’s desktop computers typically have processing power measured in gigaflops (billions of floating-point operations per second). A 30 qubit quantum computer has the potential to run at 10 teraflops (trillions of floating-point operations per second).
There is of course a problem with all this. If a particle representing a qubit is in all states at all times, how do you get a meaningful reading of its value at any one time? The answer is in something called entanglement. This effectively links two particles together so that when disturbed they will adopt opposite values, allowing the qubit to be read.
For quantum computing to be practical you need a way of controlling and reading the qubits. There are a number of possible techniques to achieve this. They can involve using magnetic or optical fields to trap ions, using light waves to control particles, dots of semiconductor material used to hold and manipulate electrons, or exploiting the impurities in semiconductors to contain particles.
Superconductors that allow electrons to flow with minimal resistance are also likely to be used in quantum computing. These need to operate at very low temperatures – absolute zero - currently, something which will need to be overcome for practical applications.
History and development
The theory of quantum computing dates from the early 1980s when it was first mooted by scientists at the Argonne National Laboratory. In 1998 researchers at MIT managed to read a single qubit in an amino acid and hydrocarbon solution.
In 2000 Los Alamos National Laboratory in the US developed a seven qubit computer in a single drop of liquid, using magnetic resonance to manipulate the particles. In the same year IBM Research developed a five qubit computer, programmed by radio pulses and read by magnetic resonance imaging similar to that used in medical scanners. IBM’s team, led by Dr Isaac Chuang, was able to solve in a single step a mathematical problem that would take repeated cycles using conventional computers.
2005 saw the University of Innsbruck create a qubyte – a series of eight qubits – using an ion trap technique. A year later scientists in the US had developed controls for a 12 qubit system. By 2007 a Canadian company D-Wave Systems was able to demonstrate a 16 qubit machine able to solve pattern matching problems like sudoku.
In 2012 the magazine Nature reported that qubits had be successfully transferred between two laboratories. In the same year Cornell University carried out the largest quantum calculation yet using 28 computational qubits.
From May of 2016 IBM has had a quantum machine available on the cloud at its TJ Watson Research Center in New York. This is currently a five qubit system but IBM’s aim is to increase its power to 50 qubits, at which point the company claims it would be capable of outperforming the world’s current top 500 supercomputers put together.
Just as the transistor replaced the valve, quantum computers look set to replace silicon chips. However, the field of quantum computing is still a very theoretical one at the moment. No one has yet managed to manipulate more than a few qubits successfully. To be a practical computing method and be able to perform real world tasks a quantum computer would need to be capable of handling at least 36 or 48 qubits. But assuming practical quantum computers can be built in the near future, what impact are they likely to have?
We’ve already seen that quantum computers are very good at things like factoring and matching patterns. This has serious implications for security. If you were to have access to a working quantum computer today, none of the encryption systems currently in use on the Internet would be safe. The flip side of this is that quantum computing could enable much more complex encryption systems that would be impossible to crack with conventional means.
A quantum computer would be able to handle much larger volumes of information and search through databases much faster than current computers too. This is good news in terms of handling the huge amounts of information likely to be generated by the Internet of Things. It could also speed up the design of machinery including, new, more powerful quantum computers.
The requirement for very low temperatures to make superconducting materials work means that a quantum computer in your home or business isn’t likely to be practical in the near future. Linking just a few quantum computers together in the cloud could lead to a quantum Internet which would make massive computing power available to everyone, and that could happen sooner than you think.