Is it fair to view this release as Nvidia strategically flexing that they can compete with their own customers in the model layer -- that they can be as vertically integrated as, say, GDM?