Fig. 17: Computational speeds and scale comparisons.
From: Coarse-grained machine learning potential for mesoscale multilayered graphene
a Computational speeds comparison for the six layers graphene between all-atom NEP (black square point), CGNEP with mapping ratio of 4:1, 16:1, and 64:1 (wine red circular points), all-atom empirical Tersoff with the LJ potential (orange inverted triangular point), TersoffCG potential with the LJ potential using parameters for CG graphene (blue upright triangular point). The mapping ratio of 1:1 denotes the potential for all-atom graphene. b Speed and size application scope of ab initio MD to CGNEP-64:1 MD calculation.
