The chemical nature and heterogeneity of most complex carbohydrates makes their structural characterization very difficult, if not impossible, through experimental structural biology. This limits our understanding of glycan-mediated recognition processes and their contribution to protein dynamics, function and shielding, all aspects of great importance in understanding viral activity. Because glycans cannot be “seen” with standard structural biology techniques, their role is often disregarded, preventing our understanding of the biological function of glycoproteins and causing delays to the development of therapies. This is concerning in view of the urgency for new approaches to detect and block viral infection against COVID-19. High-performance computing (HPC)-based molecular simulations can now provide the missing atomistic-detailed description of fully glycosylated viral envelope proteins, delivering knowledge both alternative and complementary to experiment structural biology. In this article I discuss the basic principles of biomolecular simulations, focusing primarily on glycan-specific topics and research cases concerning viral fusion glycoproteins, namely the SARS-CoV-2 S, the influenza A hemagglutinin (HA) and the HIV-1 Env trimer, where HPC provided crucial missing information about key roles of viral glycosylation.