Simulations of vector spin glasses on GPU

Spin glasses and random-field systems are models of disordered media with often highly counter-intuitive properties and an enormous number of experimental realisations in condensed matter physics, including highly frustrated magnets, superfluid helium, amorphous solids and ferroelectric materials. Novel phases of matter such as spin glasses, spin liquids and spin ices likely lie at the heart of several technologically important phenomena, such as high temperature
superconductivity, colossal magnetoresistance, and the anomalous Hall effect. Besides, models of such systems have applications in seemingly rather distant fields such as the theory of associative memory, models of the immune system or error correcting codes. These incentives notwithstanding, it was found a tricky and so-far elusive goal to understand the physics behind such systems in detail.

Although some field-theoretical advances have been made beyond the solution of the mean-field model, the analysis of short-range spin-glass systems has been largely depending on the availability of an arsenal of ever more sophisticated numerical techniques. This includes novel algorithms to cope with the extremely slow relaxation found in such systems, but also the deployment of the latest HPC systems, including special purpose computers. Based on initial implementations of the relevant algorithms on graphics processing units (GPUs), speed-ups of at least two orders of magnitude are expected for simulations of spin glasses with continuous spins as compared to conventional CPU based set-ups. In this project, a collaboration between physics and computer science is targeted at the development of a highly optimized code for the simulation of continuous-spin glasses on GPUs to finally answer some of the fundamental questions concerning the nature of the spin-glass phase in such systems.