Foxconn, the world's largest electronics contract manufacturer, reported second-quarter net profit of roughly NT$59.97 billion (about $1.86 billion), up 35% year over year and ahead of analyst expectations, driven by continued demand for AI servers. For the first time in the company's history, its cloud and networking products division, which includes AI server assembly, accounted for 51% of quarterly revenue, edging past the consumer electronics segment that includes the iPhone business Foxconn built its reputation on, which came in at 29%. The company is preparing to manufacture Nvidia's next-generation Vera Rubin server systems, with production ramp-up expected in the third quarter and shipments targeted for the fourth, and it is expanding AI-server manufacturing capacity in Mexico and Texas alongside a planned roughly 30% increase in capital spending for the year. Notably, Foxconn's own leadership flagged that next year's server output will likely be constrained less by chip supply itself and more by available CoWoS advanced-packaging capacity, which is supplied heavily by TSMC. For anyone planning capacity around GPUs, data centers, or enterprise AI infrastructure, this is a concrete signal from deep in the physical supply chain: demand for AI compute is reshaping the revenue mix and factory investment of the company that assembles a large share of the world's servers, and the next bottleneck to watch is not raw chip output but the specialized packaging step that turns chips into deployable AI hardware, which has direct implications for GPU and server lead times industry-wide.