폭스콘의 AI 서버 부문은 판매량에서 모든 기타 제품 라인을 능가하며, AI 인프라로의 결정적인 제조 전환을 시사하고 있다.
Foxconn now generates more revenue from AI servers than from the iPhone-linked operations that built its reputation, and its order book confirms that the AI infrastructure buildout represents tangible capital expenditure rather than mere earnings-call rhetoric.
Hon Hai Technology Group, the Taiwanese manufacturer globally recognized as Foxconn, has crossed a significant milestone reflecting the shifting priorities of the hardware economy. In financial results released on August 12, the company reported second-quarter revenue of NT$2.53 trillion for April through June, a 41% year-over-year increase, while net profit rose 35% to NT$60 billion. Both figures set new quarterly records.
The revenue mix shifted dramatically alongside the top-line growth. According to segment data cited by The Next Web, cloud and networking—the division encompassing AI servers—accounted for 51% of quarterly revenue, up from 48% in the first quarter. Meanwhile, smart consumer electronics, the unit most closely associated with iPhones, declined to 29%. This structural crossover marks a definitive inflection point.
The trajectory of growth is unmistakable. Reuters reported that Foxconn’s July revenue reached a record NT$946.5 billion, a 54.2% year-over-year increase, marking the first time monthly revenue surpassed NT$900 billion. The company attributed the surge to strong pull-in momentum for AI products, noting that AI rack shipments are projected to continue expanding through the third quarter.
Chairman Young Liu outlined the scale of this expansion in March, telling investors that Foxconn’s AI server rack shipments could double in 2026 while maintaining approximately 40% of the global AI server market, according to Focus Taiwan. The United States plays a central role in this strategy; the same report indicated that Foxconn’s American AI server hub is positioned to deploy 2,000 racks weekly this year. For investors assessing whether hyperscalers like Microsoft, Meta, Amazon, and Alphabet remain committed to infrastructure spending, monitoring manufacturing output provides a clearer signal than executive commentary alone.
While broader semiconductor equities recently recovered from a sharp selloff, capital flows favored memory and CPU manufacturers like Micron, Intel, and AMD over Nvidia. Micron has gained over 300% this year, compared to Nvidia’s 3.2% advance, prompting debate over whether this represents a sustained sector revival or a temporary rebound—a question explored in market analyses titled “why nvidia underperforming chip stocks rally - memory and CPU stocks outperforming nvidia.” Regardless of short-term equity rotation, Foxconn’s order book serves as a direct indicator of whether cloud providers are translating their public commitments into actual hardware procurement. When Foxconn posts strong results, it signals that capital expenditures are materializing into physical racks, motherboards, cooling systems, and contracted manufacturing capacity—a concrete measure of demand that cuts through market speculation.
Independent supply-chain data corroborates this trend. CRN reported that Dell secured $60.9 billion in AI server orders during its fiscal second quarter ended July 31, exiting the period with a $95 billion AI backlog—an increase from $51.3 billion three months prior. Dell also recognized a record $16.4 billion in AI server revenue during the same quarter. Concurrently, component pricing remains elevated. TrendForce projects conventional DRAM contract prices will rise 58% to 63% in the second quarter and another 13% to 18% in the third, while NAND Flash prices are expected to jump 70% to 75% in the second quarter and 10% to 15% in the third. Supply constraints persist across the board.
Beneath the headline revenue figures lies a critical margin disparity. Foxconn’s net profit margin on its record NT$2.53 trillion quarter stood at 2.37%, according to its own filings. By contrast, Nvidia’s latest quarterly report shows a gross margin of 75.0%. Foxconn handles assembly, soldering, and logistics, while Nvidia designs the foundational silicon that drives rack economics and captures the majority of the value. Profitability naturally concentrates at the design tier.
Foxconn does not dispute this dynamic. Operating at scale with thin margins across trillions of New Taiwan dollars still yields substantial absolute profits, and the company forecasts that demand for AI production capacity will remain robust through 2027. However, this quarter’s results highlight a more profound shift: enterprise cloud customers continue deploying capital at a pace sufficient to elevate Foxconn’s server division beyond the consumer-electronics operations that defined its legacy. While market skepticism regarding AI investment cycles will inevitably resurface, current order volumes suggest the infrastructure buildout remains firmly grounded in executed demand.