How Anthropic, OpenAI and Nvidia Are Driving the AI Economy - Forbes

As artificial intelligence continues to advance, the AI economy is entering a new phase of development. In this context, companies like Anthropic, OpenAI, and NVIDIA are playing pivotal roles, not onl... Read original →

The deep entanglement between AI model firms and chip giants is triggering a structural reshaping of the semiconductor stack. NVIDIA’s reliance on TSMC’s (Taiwan, China) 3nm EUV for H100 and Blackwell has elevated advanced packaging and optical interconnects to strategic assets, as Anthropic and peers build proprietary AI clusters. U.S. export controls, rather than stifling competition, are accelerating China’s push into chiplet and in-memory computing architectures—raising global compliance overhead. In response to CUDA’s dominance, AMD and Broadcom may target vertical-specific ASICs, while Microsoft and Google double down on custom TPUs to reduce external dependencies. Over the next 18 months, AI hardware will shift from generic compute toward co-designed model-chip systems, making control over foundry capacity and optical networking the true arbiters of AI economic power.