• Nvidia and Corning (CRNC) are deepening their collaboration to strengthen U.S.-based manufacturing for AI infrastructure, focusing on co-packaged optics and advanced fiber-optic solutions.
  • The partnership aims to improve AI network efficiency, reduce latency, and lower power usage, aligning with broader industry trends toward photonic integration.
  • This move underscores a shift in domestic supply chains for critical AI components, supported by government policies encouraging onshore production.

A strategic bet on domestic AI supply chains

Nvidia and Corning are expanding a long-term partnership to accelerate the production of advanced optics for AI infrastructure in the United States, according to people familiar with the matter. The collaboration centers on co-packaged optics and other fiber-optic technologies designed to improve the efficiency of data-center interconnects, reducing latency and power consumption in high-density AI compute environments.

Corning, a materials science giant, will supply specialty glass and fiber-optic components tailored to Nvidia’s silicon photonics ecosystem. The goal is to enable faster, more reliable data transfer between GPUs in AI factories, executives said. Both companies see this as a way to cut reliance on overseas suppliers and secure the supply chain for next-generation AI workloads.

“This is about building a domestic ecosystem for AI infrastructure,” said a Corning spokesperson, who declined to provide specific financial terms. “Our partnership with Nvidia is a key part of scaling advanced manufacturing in the U.S.”

Nvidia has been pushing to localize its supply chain, including a recent milestone with TSMC for U.S.-based wafer production. The Corning deal marks the latest step in that strategy, integrating photonics more directly with silicon compute. Industry analysts say the move could lower energy costs for AI data centers, which are grappling with surging power demands.

Market and policy tailwinds

The expanded partnership comes as the U.S. government ramps up incentives for domestic manufacturing of critical tech components. The CHIPS Act and related policies have spurred investments across the semiconductor and optics supply chain. Corning has already announced plans to add production lines at its North Carolina facility to support the Nvidia deal, according to a regulatory filing.

“We are seeing a structural shift,” said an analyst at a major investment bank, speaking on condition of anonymity. “The convergence of AI and photonics is creating a new wave of demand for U.S.-made components, and Corning is well positioned to capture that.”

Shares of Corning rose 2.3% in afternoon trading on the news, while Nvidia shares were flat. The broader market for AI data-center spending is expected to exceed $200 billion by 2027, according to industry estimates.

Broader implications

Industry observers note that the partnership could set a template for other tech companies seeking to onshore critical AI manufacturing. Similar deals have emerged among cloud providers and network equipment makers, reflecting a broader push to reduce geopolitical risks.

Still, challenges remain. Scaling up advanced optics production requires significant capital and time, and supply-chain disruptions could delay timelines. Corning has warned that its fiber-optic business faces headwinds from fluctuating demand in China.

Nvidia and Corning declined to comment further. The companies aim to begin shipping the first co-packaged optics components by early next year, according to a person close to the matter.

Correction: An earlier version of this article misstated the timeline for production ramp-up. The correct target is early next year.