The opportunity is large enough for a real compounding outcome, but not a moonshot. Rambus already ships at scale, throws off cash and sits inside AI memory bottlenecks, so it does not need a new market to appear. The upside comes from more
content per server, broader
socket coverage across memory and security, and better monetization of trusted infrastructure. The constraint is that customers and suppliers still control too many choke points for a 10x-style outcome.