In silico simulations via Langevin and Brownian stochastic dynamics have a prominent role in computational research. In the literature, there are plenty of modern algorithms providing accurate trajectories each one with its pros and cons, such as different stability regions and levels of accuracy in reproducing statistical results. The practical usability of integrators is an important aspect of each integrator: the ability to choose large time steps while ensuring numerical instability, the usage of dimensionless models, and securing numerical efficiency are at the core of stochastic integrators. In this work, several important use cases and practical features are selected in order to perform a cumulative benchmark. The effective integrators are compared. A standard industrial open-source software methodology is suggested.
Comment: 33 pages, 11 figures