Australian quantum experiment challenges computer science

Australian scientists have completed ground-breaking research using quantum computing that will challenge, among scientific principles, the theory of quantum mechanics.

A joint experiment between the University of Queensland (UQ) and Harvard University, the first of its kind to apply quantum mechanics to chemistry to predict molecular reactions, could have huge implications for science.

UQ physics professor Andrew White, a co-author of the project, said the existence of quantum computing means that either quantum mechanics is wrong, or the Church Turing Thesis, which underpins computer science, is flawed.

“If the Church Turing Thesis is wrong, that’s really big news; or it means that quantum computing will turn out to be impossible for a fundamental reason, or that a fast classical factoring algorithm exists,” White said, referring to a theory by MIT assistant professor Scott Aaronson that the only way to prove the probability of quantum mechanics is to build a quantum computer.

“If you asked the inventors of the diode what good they have done, they might have said they can shrink a computer to the size of a living room, but they would never have guessed what computers would become – this is where we are at.

“What we have done is a 2 qubit (quantum bit), toy experiment – it won’t put anyone out of a job anytime soon… but if we scale to tens and then hundreds of qubits, that’s when we will exceed the computational capacity of the planet… that will happen within 50 years.”

Due to the nature of science, the ramifications of the experiment are essentially unknown, however, White postulates that it could be used to predict the outcome of chemical reactions, albeit without the inherent randomness that is absent in controlled computer simulations.

He said it is likely that chemistry, rather than cryptography (which requires a prodigious amount of processing) will spearhead quantum computing research.

The experiment ran an algorithm dubbed the iterative phase estimation to measure the precise energy of molecular hydrogen against a predicted model. The results, White said, were "astounding" and were accurate inside of 6 parts in a million. Data was calculated to 20 bits, and in some instances up to 47 bits, and experiments were repeated 30 times for classical error correction.

Quantum computers work "brilliantly" for molecular simulations: Computational power doubles with each qubit, via the phenomena of entanglement, while the complexity of chemical reactions double with each additional atom. Simply put, no other computer, supercomputer, or bunch of supercomputers, could hope to run the simulations to the same degree.

A quantum computer with hundreds of qubits would be more powerful than every traditional computer on Earth, amounting to billions of bits. “A classical computer with 300 bits of can store 300 bits of information, whereas a 300 qubit register can store more information than the number of particles in the universe,” White said.

Scientists involved on the project included Benjamin Lanyon, Geoffrey G. Gillet, Michael E. Goggin, Marcelo P. Almeida, Benjamin J. Powell, Marco Barbieri and Harvard’s Alán Aspuru-Guzik. The experiment was funded by the Australian Research Council Federation Fellow and Centre of Excellence programs, and the US Army Research Office and Intelligence Advanced Research Projects Initiative.link