The recent implementation of the swap Monte Carlo (SWAP) algorithm bypasses computational sluggishness and gives access to the deeply supercooled liquid regime in physical dimensions. The reason for this tremendous speed-up poses an important theoretical question. Does SWAP exploits low dimensional caging defects to speed-up the dynamics or is it efficient to break robust cages further in high dimensions too? We systematically study the efficiency of SWAP in high dimensions and show that it decays rapidly with increasing dimension at fixed sluggishness. Interestingly, SWAP nonetheless delays systematically the putative dynamical transition by an amount that seemingly survives in the limit $d\to \infty$. Our results open the possibility of studying finite-dimensional deviations from the mean-field description of glasses by accessing pure glassy dynamics away from the regime in which local structures and activated processes interfere.

Bypassing glassy sluggishness by particle swap Monte Carlo in high dimensions

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.