US-China AI Summit Focuses on Compute Supply Chain Control

The upcoming Trump-Xi summit marks a critical inflection point, shifting the U.S.-China AI rivalry from a broad-based trade conflict to a targeted battle over the physical supply chain for AI computation. This isn't just about tariffs; it's a strategic renegotiation of global technology access, with AI hardware and its constituent rare earth minerals now serving as the primary bargaining chips. Whereas previous disputes centered on IP theft and market access, this summit frames the entire AI ecosystem, from TSMC's foundry capacity to China's dominance in gallium and germanium processing, as a zero-sum contest for national security and economic dominance. The mechanics of this negotiation expose the deep interdependencies and vulnerabilities of both nations. The U.S. wields its most potent weapon: control over high-end semiconductor designs and manufacturing equipment, directly threatening China's ability to produce advanced AI accelerators. In response, China leverages its near-monopoly on rare earth mineral processing, which is critical for everything from NVIDIA's GPUs to U.S. defense systems. A key loser in any escalation would be non-Chinese Asian economies like South Korea and Taiwan, whose semiconductor industries would be squeezed between U.S. export controls and Chinese resource restrictions. Looking forward, the summit's outcome will determine the next decade's technology landscape. A deal might establish a bifurcated tech world with explicit "red lines," while a breakdown could trigger an all-out supply chain war, forcing immediate reshoring efforts and massively inflating hardware costs. The critical variable will be whether both sides can accept a managed rivalry over an unconstrained one. The real test is not the rhetoric, but the specific export licenses granted or withheld by the Commerce Department and China’s Ministry of Commerce in the three months following the summit.