Generation of well-converged conformational ensembles have remained a major challenge in atomistic simulation of proteins and other biomolecules. This is particularly important for studying so-called intrinsically disordered proteins, which rely on a lack of stable structure for function and are enriched in cancer and other disease pathways. I will describe a multi-scale enhanced sampling (MSES) method where efficient topology-based coarse-grained models are coupled with all-atom ones to enhance the sampling of atomistic protein energy landscape. The bias from the coupling is removed by Hamiltonian replica exchange, thus allowing one to benefit simultaneously from faster transitions of coarse- grained modeling and accuracy of atomistic force fields. The method is demonstrated by calculating the conformational equilibria of several small but nontrivial β-hairpins with varied stabilities. We anticipate MSES to be highly useful whenever generation of well-converged protein conformational ensembles is critical, including intensive current efforts that rely on peptide simulations to optimize implicit and explicit solvent protein force fields.