Nvidia H200

Nvidia H200 is the Nvidia AI accelerator discussed in Bytes: Week in Review - New chip exports for China, Microsoft to pay electricity for AI data centers, and Gemini will power Apple’s AI as newly eligible for sales to China under a security-rule and revenue-sharing arrangement. The Marketplace Tech Bytes episode says the U.S. government would receive 25% of sales and that enforcement remains the central question.

The page matters to the wiki because H200 turns AI Export Controls from an abstract hardware-versus-model-access contrast into a transactional policy case. In the source’s framing, U.S. officials try to preserve control and revenue, Jensen Huang argues for keeping Chinese AI infrastructure tied to American chips, and China has incentives to support alternatives such as Huawei through Domestic AI Chip Catch-Up.

Connections

• Nvidia and Jensen Huang - company and executive context.
• AI Export Controls - policy frame for controlled access to advanced AI chips.
• China, Huawei, and Domestic AI Chip Catch-Up - substitution and industrial-policy branch.
• Strategic AI Infrastructure Dependence - infrastructure dependence shaped by chips, policy, and customer access.