Ab initio quantum chemistry seeks to describe and elucidate chemical species and processes using quantum mechanics. For dynamical chemical processes, molecular dynamics (MD), where the atoms of a chemical system move according to Newton's laws of motion, is frequently used. MD calculations have historically used classical mechanics rather than quantum mechanics to describe the evolution of a chemical system. The use of classical mechanics with MD has proven to be a great success, but classical MD has deficiencies, since quantum mechanics must be used to describe important chemical phenomena such as bond breaking or excited states accurately. With the increase of computer power over the past half-century, ab initio MD (AIMD) methods that describe a chemical system using quantum mechanics have been developed to eliminate the deficiencies of classical MD.

Unfortunately, the application of AIMD is limited to small systems and short time scales since standard quantum chemical methods exhibit non-linear scaling with system size. More recently, new approaches have circumvented the non-linear scaling of quantum chemical methods by exploiting the fact that most chemical interactions are local and therefore distant interactions can be approximated or even ignored. Other methods obtain quantum mechanical accuracy at a cost associated with classical mechanics by deriving a classical force field directly from ab initio calculations. Individually and in combination, methods that eliminate the non-linear scaling of standard ab initio methods have the potential to extend the reach of AIMD to larger systems such as surfaces, molecular clusters, bulk liquids, and proteins.