This story is over 5 years old.

Silicon Quantum Logic Gates Offer a Glimpse of Practical Quantum Computing

Common manufacturing materials for an uncommon technology.
Image: UNSW

For the first time, physicists have crafted a two-qubit quantum logic gate using a common computer manufacturing material: silicon. It's a major, if disconcertingly hyped, advance.

In a paper published this week in Nature, a University of New South Wales-based research group describes a logic gate based on the correlated spin states of electrons. The result is a system that "provides all of the necessary operations for universal quantum computation," according to the researchers. The group has patented their work and is now seeking industry partners to do the engineering—or rework some existing engineering—and make quantum computing a reality.


The usage of quantum dots in quantum computing systems is not a new idea. In fact, it's among the most intuitive. The dots, which function to confine the motions of electrons, can be manufactured in large blocks in a way that resembles the arrangements of cells in a standard computer memory module. Once trapped, the electrons can be manipulated in different ways using electrical pulses or magnetic fields such that they exist in defined states representing information as 1s and 0s.

As the UNSW group explains, using quantum dots is a nice idea, but trapped particles tend to "dephase" thanks to background forces arising from the spins of nearby atomic nuclei. That's where silicon comes in. Implementing quantum dots using purified forms of the material offers spin states with fidelities above the threshold of the quantum computer's error handling abilities.

Implementing quantum dots in silicon is not a new thought, but to make it actually count, engineers need to be able to actually do stuff to and with those dots. This is the significance of the UNSW work. "A scalable approach towards quantum computation ideally requires that the coupling between qubits can be turned on and off, so that single- and two-qubit operations can be selectively chosen," the group writes. This selective coupling is what they actually accomplished.

The basic idea is to take two individual qubits, which are particles representing information as probabilistic combinations of 1s and 0s (rather than deterministic 1s or 0s), and manipulate them into the desired states using an oscillating magnetic field. The particles are then coupled and decoupled using electrical pulses. The gate is represented by the exchange interactions between the two qubits—as a quantum CNOT gate it works by entangling and disentangling particles from one another, which is more or less the central feature of a quantum computer.

"If quantum computers are to become a reality, the ability to conduct one- and two-qubit calculations are essential," said Andrew Dzurak, the lead author of the new study, in a statement. "We've morphed those silicon transistors into quantum bits by ensuring that each has only one electron associated with it. We then store the binary code of 0 or 1 on the 'spin' of the electron, which is associated with the electron's tiny magnetic field."

This is all pretty cool, but there are a lot of different approaches to building a quantum computer using all sorts of mechanics: trapped ions, Bose-Einstein condensates, nuclear spins, optical lattices, and superconducting circuits. That's just to start. In August, for example, I wrote about quantum logic gates based on light particles, what are often assumed to be the information carriers of the quantum future, rather than electrons. Whether the QC future is in fact carved into silicon remains to be seen.