Cristian Micheletti
(SISSA)
 

Sampling equilibrium ensembles of dense polymer mixtures is a paradigmatically hard problem in computational physics, even in lattice-based models. For instance, using real-space Monte Carlo to sample polymer systems becomes impractical for increasing size, rigidity, and density of the chains. In response to these challenges, we introduce and apply a formalism to recast polymer sampling as a quadratic unconstrained binary optimization (QUBO) problem ^[1].
Thanks to this mapping, dense systems of stiff polymers on a lattice can be efficiently sampled with classical QUBO solvers, resulting in more favourable performance scaling compared to real-space Monte Carlo ^[2]. Tackling the same problems with the D-Wave quantum annealer leads to further performance improvements ^[2]. As an application, we discuss the use of the quantum-inspired encoding on a hitherto untackled problem, namely the linking probability of equilibrated melts of ring polymers, for which we unveil counterintuitive topological effects.
 

Zoom registration link:

https://zoom.us/meeting/register/tJckcu6oqDsrH9EyjOc2PpUD3J3pR08YnHHc