Efficient parallel algorithms for Monte Carlo simulations of millions of water molecules in the fluid phase

Simulating water droplets made up of millions of molecules and on timescales as needed in biological and technological applications is challenging due to the difficulty of balancing accuracy with computational capabilities. Most detailed descriptions, such as ab initio , polarizable, or rigid models, are typically constrained to a few hundred (for ab initio ) or thousands of molecules (for rigid models). Recent machine learning approaches allow for the simulation of up to 4 million molecules with ab initio accuracy but only for tens of nanoseconds, even if parallelized across hundreds of GPUs. In contrast, coarse-grained models permit simulations on a larger scale but at the expense of accuracy or transferability. Here, we consider the CVF molecular model of fluid water, which bridges the gap between accuracy and efficiency for free-energy and thermodynamic quantities due to i) a detailed calculation of the hydrogen bond contributions at the molecular level, including cooperative effects, and ii) coarse-graining of the translational and rotational degrees of freedom of the molecules. The CVF model can reproduce the experimental equation of state and fluctuations of fluid water across a temperature range of 60° around ambient temperature and from 0 to 50 MPa. In this work, we describe efficient parallel Monte Carlo algorithms executed on GPUs using CUDA, tailored explicitly for the CVF model. We benchmark accessible sizes of 17 million molecules with the Metropolis and 2 million with the Swendsen-Wang Monte Carlo algorithm.