05-10-2018 | By Rob Coppinger
Quantum computers could be mass produced at low cost with silicon microchips that have large-scale waveguides, optical tracks, for photons, instead of circuits for electrons controlled by transistor switches.
These silicon photonic chips could solve two challenges for quantum computing, low-cost, high quality production methods and creating a reprogrammable computer. Silicon chips’ manufacturing methods are well proven and chips with large-scale optical waveguide tracks will enable mass production of quantum processors. The silicon quantum processor works by guiding its photons, packets of electromagnetic energy, along its large-scale waveguides and encodes them into quantum-bits of information called qubits. A qubit can be a one and a zero at the same time and it is this superposition quality that allows the simultaneous calculations that can outperform conventional computing.
Artist’s illustration of the chip. Tracks called waveguides guide photons in silicon, much like an optical fibre. Spirals of these waveguides are used to generate photons (quantum particles of light) that are then routed around the processor circuit to perform different tasks. Credit: Xiaogang Qiang / University of Bristol
“It’s a very primitive processor [our silicon photonic microchip], because it only works on two qubits, which means there is still a long way before we can do useful computations with this technology,” explains Xiaogang Qiang, who undertook the research while a doctoral student at the University of Bristol. He now works at the National University of Defence Technology in China. “What we’ve demonstrated is a programmable machine that can do lots of different tasks,” he adds.
Any task that can be achieved with two qubits, can be programmed and realised with the processor. This small device has more than 200 photonic components and can be used as a scientific tool for quantum information experiments to begin with. It was made with Complementary Metal Oxide Semiconductor, CMOS, compatible processes and the Bristol team programmed it to implement 98 different two-qubit operations. Qiang’s former colleague, Dr. Jonathan Matthews, said: “We’ve used this device to implement several different quantum information experiments using nearly 100,000 different re-programmed settings.”
Future photonic microchips with more qubits will be able to take advantage of quantum particles’ unique property of entanglement. This property links qubits, entangling them, at any distance, and this linking enables the simultaneous computing that makes these quantum machines faster than conventional electronics. For example, two qubits in superposition are entangled and together they can store all the possible combinations of their quantum states, resulting in four values; two ones and two zeros. As qubits are added, the number of combinations increases dramatically, for example, 20 entangled qubits can store more than a million values.
Matthews added: “We need to be looking at how to make quantum computers out of technology that is scalable, which includes technology that we know can be built incredibly precisely on a tremendous scale.” Matthews works in the University’s Quantum Engineering Technology Labs, which was launched in April 2015, encompassing the science and engineering faculties with more than 100 researchers and students, 12 core academics and 40 associated academics. The quantum processor work at Bristol began in 2008 to investigate alternatives to the methods using mirrors, lasers and optical elements to create quantum circuits.
Read more electronics news related to advancements in quantum computing: Quantum cascade lasers could help transistors become more energy efficient