The outstanding challenges in computer simulations of biological macromolecules are related to their complexity. Part of the complexity of biological systems concerns their physical size. Enumerating atoms ranging from a few in small signal molecules to the millions of particles in biological complexes is an obvious example of biological hierarchy. Another aspect is the extremely broad range of timescales of life science processes (many orders of magnitude); this adds another dimension of complexity. This extended range of timescales may even be observed for a single biomolecular process. Consider, for example, the R to T transition in hemoglobin. The complete conformational change occurs in tens of microseconds. However, the system has more than one timescale. Considerable activity occurs on a range of timescales before the final event (heme relaxation, picoseconds; tertiary relaxation, nanoseconds; ligand escape from the protein matrix and rebinding, hundreds of nanoseconds and so on). Whereas the basic time-step of atomically detailed simulations is about a femtosecond, it is not difficult to find molecular processes in biology that span more than ten orders of magnitude of relevant times, making the straightforward simulation of these events very difficult. Several techniques have been developed in recent years to address these problems.