UK scientists claim first calculation on optical quantum chip

University of Bristol team uses photonic chip for primitive cryptographic calculation

Optical quantum computing first for University of Bristol

A big step towards construction of a massively powerful quantum computer has taken place, following research by a team of UK-based physicists and engineers.

The experiment was performed by PhD students Alberto Politi and Jonathan Matthews, currently doing research at University of Bristol's Centre for Quantum Photonics.

The mathematical computation investigated ways to find the prime factors of 15, which although trivial, points the way to more important quantum-based calculations, once the hardware can be scaled up to herald fully functional optical quantum compute systems. The system outputs the correct prime factors of 15 as three and five.

The three main components of the University of Bristol's system was a photon source, a single photon detector, and circuitry sandwiched in between to perform a simple cryptographic calculation.

Matthews said that essentially, a compiled version of Shor’s factoring algorithm on a chip was being demonstrated.

"Factoring is hard if you use classical computation techniques, since as the size of the number scales, the calculation time scales exponentially, whereas with Shor’s algorithm the calculation time scales only polynomially," he said.

Easy factorisation of large prime numbers using quantum compute systems could render current encryption techniques used over the internet useless.

When a proper quantum computer exists, added Matthews, "it will actually be working faster on a fundamental level than any classical computer could do, and that’s one of our main points."

Asked if the state of quantum computers was at a similar stage to the early transistor-based classical computer systems, Matthews said: “That’s a fair analogy."

Matthews explained that world-renowned physicist Richard Feynman suggested that if you wanted a fully accurate simulation using quantum systems, for example of a chemical molecule on a molecular level, there may be some benefit to calculations which are themselves based on quantum mechanics.

"It is quite difficult to say when we’ll have fully scaleable quantum computers – it could be of the order of 30 to 50 years – that type of timescale," he said.

"There are still a lot of challenges to get over and you couldn’t see factorisation of large numbers within the next 10 years."

That estimate of 10 years means that current encryption schemes used over the internet are safe, and would not be easily breakable for a long time.

Politi et al's paper, Shor’s Quantum Factoring Algorithm on a Photonic Chip, was published in Science yesterday.