While the relevant features and properties of nanosystems necessarily depend on nanoscopic details, their performance resides in the macroscopic world. To rationally develop and accurately predict performance of these systems we must tackle problems where multiple length and time scales are coupled. Rather than forcing a single modeling approach to predict an event it was not designed for, a new paradigm must be employed: multiscale modeling. A brilliant solution to a pervasive problem, Multiscale Modeling: From Atoms to Devices offers a number of approaches for which more than one...