Canadian researchers think they may be bringing the industry one step closer to the construction of powerful computers that use light instead of silicon.
Today’s computer memory consists of a string of 0s and 1s, and even the largest supercomputers have to handle them one at a time. Quantum computers, as they are known, would have memory capable of processing calculations of several numbers at once.
Those specializing in this field are trying to find ways that the smallest units of light (photons) could replicate the way transistors in a PC control electrons. Transmitting data this way is extremely difficult, but Aephraim Steinberg, a physics professor at the University of Toronto, says he and his students may have a way of simplifying the process.
Steinberg and his research team are using beams of light and optical crystals to manipulate the photons. When the researchers shine a strong beam of light and two weak beams of light on an optical crystal, for example, they collide and turn each other off. This is the “switch” that would be the basis of an optical transistor — the building blocks for quantum computers.
“As soon as there’s any record of whatever the computer is thinking about, that destroys quantum computing,” Steinberg says. “If a transistor gets a little hotter thinking about a 1 than when it’s thinking about a 0, that heat is enough to destroy the quantum power of the computation. You need to find a system where nothing gets sent to the outside world about what it’s doing.”
Photons are considered ideal ways to store quantum information because they are very isolated from the outside world. As it travels through a fibre, air, or through a piece of glass, light generally moves undisturbed without giving off heat.
To make a computer, you have to not only store data but build a “logic gate” to make the data interactive. Since photons or light beams travel through each other without colliding, it seemed like the industry would never be able to build a switch or a transistor for single photons. “We were in this situation where atoms and electrons had really nice switches, but so much noise that they couldn’t behave quantum mechanically,” he says. “Light behaved quantum mechanically but no one knew how to build a switch for them.”
Steinberg says the switch he and his team have developed is not exactly the one quantum theorists have been looking for, but it may be applicable towards the development of the machines. There is already considerable interest among companies like IBM and HP that serve high-performance computing environments in furthering this effort. In Vancouver, a company called D-Wave Systems Inc. is specializing in the creation of quantum “bits” that would replace the traditional bits we use in today’s computers.
Dr. Geordie Rose, D-Wave’s president, says quantum computing research is making big strides towards commercial availability.
“People usually underestimate how quickly new technologies are developed if there’s a market pull,” he says. “If there is a definite need for something to happen, innovations happen quicker than people imagine they can.”
In quantum computing, the market pull is coming from the bioscience companies, he says. “Those people have problems which have complexities that are not tractable in our lifetimes using silicon-based technologies,” he says. “They’re looking for some solution to their heartache, and one of the potential solutions are quantum computers.”
Steinberg isn’t as sure.
“I’m a bit of a pessimist – I think there’s a 50 per cent chance that we’ll just never build a working quantum computer,” he says. “If we do, my guess is that we’ll have them in less than 10 years. A lot of the ideas are there — it’s just a question of figuring out which system is right and perfecting some techniques.”
Even if quantum computers are never realized, Steinberg adds that the industry may learn of ways to improve heat reduction and create better versions of silicon-based machines.