158 …Custom chips also reflect a broader effort to manage the economics of AI infrastructure. As Amazon CEO Andy Jassy noted in early 2025, AI does not have to be as expensive as it is today, and it won’t be in the future. Custom silicon is one lever to control these expenses. At the same time, a new ecosystem of infrastructure specialists is emerging to meet this demand. CoreWeave has become one of the fastest-scaling cloud GPU providers, repurposing gaming and Crypto hardware supply chains to serve enterprise AI customers. Oracle, long seen as a legacy IT vendor, has repositioned itself as a GPU-rich cloud platform with AI-specific offerings. Astera Labs, a lesser-known but critical player, builds high-speed interconnects that move data between GPUs and memory systems with minimal latency – an increasingly important performance constraint. These firms aren’t building foundation models, but they’re building what foundation models depend on. As compute demand compounds, they’re becoming essential infrastructure in a market where speed, availability, and efficiency are important differentiators. AI-Related Monetization = Very Robust Ramps