We want to further eliminate friction among heterogeneous architectures and applications without compromising their distinctive underlying capabilities.