Row upon row of microscopic hockey pucks etched into a silicon chip could hold the key to making quantum computers at practical scale and allow for un-crackable encrypted information to be sent over long distances.
So says a Canadian research team that developed the chip as an alternative way to create a quantum computer.
Quantum computers are different from conventional computer systems, in which bits of information are stored electronically as ones or zeros. In a quantum computer, the bits are replaced with qubits, each of which carries some probability of being either a one or a zero before it is measured. This ambiguity is what gives a quantum computer its power to perform certain types of calculations far more rapidly than a conventional system.
One example is mathematical operations involving large prime numbers, the basis for keeping much of the world’s data secure.
In some quantum computers, qubits take the form of electric circuits in a supercooled material. In others, they are made up of particles of light moving through a series of optical devices. Each approach comes with its own drawbacks.
Stephanie Simmons and colleagues at Simon Fraser University in Burnaby, B.C., have aimed at another method that places a quantum computer on a silicon chip – a material that has long been the workhorse of conventional digital technology and easy to mass produce.
In their approach, each qubit resides in a tiny flaw in the atomic structure of the chip, where a single atom of silicon is substituted with two carbon atoms and one atom of hydrogen. Such a structure can exhibit a quantum property called spin, which can be measured as either up or down. By matching each direction to a value, this spin property then becomes the one or zero of the qubit.
In a report published last week in the journal, Nature, team members demonstrated that they had created and identified 150,000 such qubits on a single chip in a way that ensured each qubit could be individually accessed and controlled. The hockey puck structures they etched into the silicon contain, on average, one qubit each.
Each disk is “slightly taller, but still pretty close to the proportions of a hockey puck – just 100,000 times smaller,” said Daniel Higginbottom, lead author on the report.
Quantum computers that work at commercial scale are expected to require millions of qubits. Dr. Simmons said the approach she and her colleagues have taken can be extended to that scale. The qubits can also be directed and connected with each other by using light passed through fibre optics – a well-developed technology that already underpins modern communications hardware.
If successful, the approach would provide a convenient way to move quantum information from qubits that are located in a solid into a light-based system which can send it over a long distance. This would be a pre-requisite for developing a new style of encryption system that cannot be cracked by quantum computers in the future.
“It’s a really slick kind of solution to what people have been looking for, and that’s why we’re so excited,” Dr. Simmons said.
Aephraim Steinberg, a University of Toronto physicist who was not involved with the work, said that the team had cleared an important hurdle in demonstrating the value of their approach, though it remains to be seen which kind of quantum system will ultimately prove to be the most useful.
Dr. Steinberg, who directs a program on quantum information science for the research organization CIFAR, added that a key feature of the silicon-based system is the way it connects a form of quantum technology that is well-suited for memory and processing with another that is well-suited for communication.
“Since we’re going to need to do both … we need to learn how to interface our memory and processing chips with the fibres that are going to transmit the information,” he said.