Boosting the accuracy and speed of quantum Monte Carlo: Andrea’s paper published in PRB

Andrea and Dario’s paper on the methodological improvement of the diffusion quantum Monte Carlo algorithm is finally published in Phys. Rev. B. In this work, involving also Sandro Sorella and Michael J. Gillan, we make the method up to one hundred times faster, by modifying only one line of the CASINO code. We also solve an outstanding issue of diffusion Monte Carlo, related to size-consistency, that was hampering its accuracy in many practical calculations.
It is now possible to study a whole variety of systems, including calculating interaction energies between molecules or between molecules and surfaces, or investigate the relative stability of different crystal polymorphs, with unprecedented accuracy and at a much lower computational cost.
The new method extends the range of the size of the systems that it is now possible to simulate, opening new avenues to the study of previously unaccessible problems in Nanotechnology and Material Sciences.

Group gatherings