Bright ideas for the future of computing

NTU scientists have secured competitive funding from Singapore’s Ministry of Education to harness light at quantum scales for simulating and probing molecular structures.

When you stare at a light bulb, billions of photons—particles of light—released by the bulb travel into your eyes. But travelling light is much more interesting when photons go solo; they take on unusual properties that have far-reaching consequences for the field of computing.

In existing computers, the foundational unit of data is a binary digit, or bit, so called because it can take the value of either 0 or 1, but not both. But if this data were to be encoded using particles such as photons, you instead get a quantum bit—or qubit—which can take the value of both 0 and 1 at the same time. This property is known as superposition. The qubit multitasks, if you like, thereby allowing a quantum computer to be significantly faster and more powerful than a conventional one.

“Quantum computing is an extremely hot research area right now. It has the potential to surpass even supercomputers in terms of speed and computational volume, and could completely change the way the world processes information,” says Prof Liu Ai Qun of NTU’s School of Electrical and Electronic Engineering.

However, to date, practical integrated circuits for light-based computing remain elusive. Prof Liu and his colleagues intend to change that with funding of up to S$10 million (US$7.3 million) over five years recently secured from Singapore’s Ministry of Education for a project called “Quantum computing chip and 3D nanophotonic integration”.

Riding the quantum wave
“We want to create a silicon-based photonic chip with waveguides on it—tiny fibres, each with a silicon core and an oxide cladding—that can control the path of individual photons. This will let us harness quantum effects for computing,” says Prof Liu. To make this chip a reality, he has teamed up with Assoc Prof Kwek Leong Chuan of NTU’s National Institute of Education, who is also Co-Director of the Quantum Engineering Programme under Singapore’s National Research Foundation and a principal investigator at the Centre for Quantum Technologies hosted by the National University of Singapore.

“Each waveguide would be thinner than a human hair—approximately 420 nanometres wide and 220 nanometres thick. Assuming we have just two waveguides, we could designate them as a 0-path and a 1-path. As photons travel along these two paths, you get superposition and interference among the photons, which enable rapid computation,” Assoc Prof Kwek explains.

Prof Liu adds that the maximum computing power of this platform can be compounded exponentially by increasing the number of waveguides and photons on a chip. Hence, his target is to manufacture a chip with 50 paths that can manipulate 20 photons simultaneously.

“Such a chip would achieve quantum supremacy, which means that we would be able to perform certain calculations beyond the capacities of classical computers,” he says.

Finding chemistry with computing
With great computing power comes a great number of possibilities. Prof Liu and Assoc Prof Kwek have decided to use their proposed quantum photonic chip to solve the problem of simulating chemical structures.

Because photons can mimic atoms and electrons in devices and materials, photonic chips can harness the quantum features of photons to perform calculations in quantum chemistry faster and more efficiently than classical computers. This applies particularly to problems involving complex molecules that comprise multiple atoms, which quickly become intractable even for supercomputers. The speedup in computation could ultimately lead to better design and discovery of new pharmaceutical products, materials and industrial catalysts.

“Fortunately, the structures and interactions of some atoms are already known and mathematically described, so we ought to be able to design our waveguides to simulate those atoms on a chip,” Assoc Prof Kwek notes.

Eventually, with a repository of waveguide configurations representing known chemical structures, one could hypothetically mix and match those configurations to simulate larger and more complex molecules. Conversely, given the light-scattering spectrum of an unknown compound (or even a mixture of unknown compounds), one could reference this database of simulated molecules to learn its identity.

“Simulations of chemical structures could pave the way for novel materials with unique and useful properties, as well as facilitate the development of better sensors. Our proposed quantum photonic computing chip could therefore accelerate progress in a variety of scientific domains,” says Prof Liu.

A superposition of strengths
Although competition is stiff among labs seeking to develop the world’s first quantum photonic computer, Prof Liu and Assoc Prof Kwek are quietly confident that they can leapfrog the bigger players in the field.

“Singapore has built up strong technical capabilities in the semiconductor industry over the years, so we are well positioned to take the lead in silicon-based fabrication of quantum computing chip technology,” says Prof Liu.

Acknowledging that breakthroughs are not achieved in silos, both researchers also point out the multidisciplinary nature of their teams. “NTU is a powerhouse in engineering research internationally, but what’s important is that our engineers work closely alongside physicists so that theory and applied science complement each other,” Prof Liu states, adding that this collaborative approach extends beyond Singapore’s borders to include overseas partners who “help us learn and advance very quickly”.

Wasting no time, Prof Liu and Assoc Prof Kwek have already embarked on creating a prototype based on their ideas. They hope to develop a functional, validated chip for chemical structure simulation within five years.

“Quantum computing will, without a doubt, be a transformative technology. Countries that don’t build capability in this area run the risk of being left behind when their classical computing systems become obsolete,” Prof Liu concludes.

You may also like...