Comment on 'Computation of the viscosity of a liquid from time averages of stress fluctuations'

In a recent paper, Hess and Evans [Phys. Rev. E 64, 011207 (2001)] propose a method different from the conventional Green-Kubo and Einstein methods to calculate viscosity in equilibrium molecular dynamics simulations. For a comparison, we calculate the shear viscosity of SPC/E water at 303 K using these three different methods. We find that the Hess-Evans method is not as good as the other two in practical application, especially for the fluids with high viscosity and complicated relaxation.