Foxconn’s Post-Apple Pivot: AI Servers Now Lead Its Growth Engine

Taipei-based Foxconn, long synonymous with Apple assembly, has diversified decisively into AI servers as smartphone demand matured. In Q2, consumer electronics accounted for 35% of revenue versus 41% from cloud and networking, a reversal from 2021 when handsets dominated the mix. Analysts credit the change to strategic bets taken well before the current AI wave, reducing the risk of over-reliance on iPhone cycles and stabilizing top-line momentum as data-center spending accelerates worldwide.

The company’s deepening relationship with Nvidia is central. Foxconn began producing reference designs for Nvidia graphics cards in the early 2000s and general-purpose servers around 2009, giving it a head start as AI training and inference workloads exploded. It is now one of the world’s largest suppliers of both general-purpose and AI servers and plans new capacity in Houston and Mexico to serve U.S. clients. Management guides for AI-server revenue to grow more than 170% year-on-year in Q3, underscoring how quickly this line is scaling.

This pivot is industry-wide across Taiwan’s original design manufacturers. Wistron’s January–July revenue rose 92.7% and Quanta’s 65.6%, reflecting stronger server demand from U.S. tech giants. Market researchers estimate Taiwan now ships about 80% of global servers and more than 90% of AI servers, a dominance built on close, long-running collaboration with cloud providers and on rapid, modular manufacturing. The result is a structural reweighting of the island’s tech sector toward enterprise and AI infrastructure rather than consumer gadgets.

While Foxconn’s moves into EVs and chips have yet to meaningfully lift sales, its AI-server momentum shows the pay-off from earlier risk-taking and vertical integration. The company’s willingness to invest ahead of demand and to diversify production footprints has strengthened customer ties and insulated revenue from handset slowdowns. If data-center capex remains firm, Foxconn and its Taiwanese peers are positioned to capture the next leg of AI infrastructure build-out.