Molecular dynamics simulations are useful tools to unveil molecular mechanisms of polymer phase separation, self-assembly, adsorption, and so on. Due to large molecular size and slow relaxation of the polymer chains, a great amount of issues related to large-distance chain displacement cannot be tackled easily with conventional molecular dynamic simulations. Systematic coarse-graining and enhanced sampling methods are two types of improvements that can boost spatiotemporal scales in polymer simulations. We present two typical ways to obtain the coarse-graining potential either by fitting to correct liquid structures or by fitting to available thermodynamic properties of polymer systems. The newly proposed anisotropic coarse-grained particle model can be used to describe aggregation and assembly of polymeric building blocks from disk-like micelles to Janus particles. We also present a stochastic polymerization model combined with coarse-grained simulations to investigate the problems strongly influenced by the coupling of polymerization and excluded volume effects. Finally, a facile implementation of integrated tempering sampling method is illustrated to be very efficient on bypassing local energy minima and having access to true equilibrium polymer structures. Molecular dynamics (MD) is a powerful simulation technique proven to produce highly realistic results in a wide variety of applications over the past decade [1,2]. However, the computational costs of detailed interaction models in this paradigm severely limit its applicability beyond extremely small spatiotemporal scales. In fields such as polymer and molecular biology, many interesting phenomena occur in length and time scales much bigger than those pertaining to the motion of single atom or molecule. Hence both coarse-grained MD and dissipative particle dynamics (DPD) techniques [3,4] were extensively developed and used in order to allow simulations of complex soft matter systems in larger scales more relevant to these processes [5][6][7][8][9], by omitting degrees of freedom not immediately essential for the description of systems at studied level [10][11][12][13]. For slowly-evolving polymer systems, it is also of great interest to resort to enhanced sampling techniques developed recently for approaching better thermodynamic average calculations [14][15][16][17]. Therefore, by combining both systematic coarse-graining and enhanced sampling methods, it is possible to investigate polymer systems with extended spatiotemporal scales that are much relevant in experiments but inaccessible with conventional atomistic detailed MD simulations.