NVIDIA And Corning: Long-Term Partnership Expands U.S. AI Infrastructure Manufacturing - Pulse 2.0

NVIDIA and Corning have entered a multiyear commercial and technology partnership aimed at expanding U.S.-based manufacturing capacity for optical connectivity products essential to AI infrastructure.... Read original →

NVIDIA’s strategic lock-in with Corning signals a pivotal shift in AI infrastructure—from GPU-centric scaling to full-stack physical-layer reengineering. Technically, the insatiable bandwidth demands of dense GPU clusters elevate optical interconnects from ancillary components to critical bottlenecks; Corning’s tenfold U.S. capacity expansion will accelerate adoption of multimode fiber and silicon photonics at 800G/1.6T speeds, pressuring rivals like Coherent and II-VI to integrate faster. From a compliance standpoint, this move exploits CHIPS Act incentives to onshore sensitive optical manufacturing, preempting potential export controls. Competitively, AMD-Lumentum and Broadcom-Inphi alliances may be forced into deeper integration. Within 18 months, a U.S.-centric ‘compute-optics-manufacturing’ triad will solidify—but at the risk of inflating global deployment costs and triggering parallel non-U.S. supply chains.