Simulating Physics with Computers: a Myth?
Professor Peter Drummond
Centre for Atom Optics and Ultrafast Spectroscopy, Swinburne University of Technology
3:30 pm Friday, 9 November 2012, EN313, EN Building, Hawthorn.
Feynman's 1982 paper, entitled `Simulating physics with computers' was a pioneering study in quantum simulation. He asked the central question: `Can quantum systems be probabilistically simulated by a classical computer?', and answered it himself by stating: `If ..
there's no hocuspocus, the answer is certainly, No!'
Feynman supported this conjecture by appealing to Bell's theorem, which rules out classical hidden variable theories equivalent to quantum mechanics.
Explicit counterexamples are given to this wellknown claim, through direct probabilistic simulation of quantum correlations violating Bell's inequality. Bell violations are also demonstrated in highly nonclassical multipartite states. These have up to fifty qubits, corresponding to a basis of a quadrillion quantum states.
Feynman's concerns can indeed be overcome. Probabilistic simulations of Bell violations can certainly be carried out, although very challenging in the multipartite case. These types of multipartite Bell measurements are potentially accessible in experiments on iontraps and similar technologies, closely related to those studied by the 2012 Nobel Laureates.
Back to 2012 programme
