Are pressure fluctuation-based equilibrium methods really worse than nonequilibrium methods for calculating viscosities?

Nonequilibrium methods have become increasingly popular for calculating the viscosity of liquids. The use of nonequilibrium methods is frequently rationalized by the notion that pressure fluctuation-based equilibrium methods, such as those using the Green–Kubo GK formula and the Einstein relation, do not converge well, and hence are less suitable than nonequilibrium methods. In this short note, we revisit the systems on which this conclusion is based. We show that a slightly different way of analyzing the data allows us to use equilibrium molecular dynamics methods to compute viscosities with comparable accuracy and reliability as the nonequilibrium techniques considered in Ref. 1 and the reverse nonequilibrium molecular dynamics method. The shear viscosity can be computed using the GK formula