Foxconn Reports Record Q2 Results as AI Server Demand Drives 2026 Growth

Hon Hai Technology Group (Foxconn) reported record second-quarter revenue, operating profit and net profit as demand for AI servers continued to drive growth across its cloud and networking business.

Revenue for the April-June quarter reached NT$2.53 trillion, up 41% year over year, according to the company’s Q2 2026 financial results. Operating profit increased 68% to NT$94.8 billion, while net profit rose 35% to NT$60 billion. Earnings per share reached NT$4.27.

Foxconn also reported NT$4.65 trillion in revenue for the first half of 2026, a 35% increase from the same period a year earlier. Operating profit rose 65% to NT$170.5 billion, while net profit increased 27% to NT$109.9 billion.

The results come as AI infrastructure investment continues to reshape demand for the manufacturers and suppliers building data center systems.

AI Racks Become a Larger Growth Driver

Foxconn Rotating CEO Michael Chiang said the company continued to see volume growth in AI server racks as cloud service providers expanded capital expenditure.

Cloud and networking revenue is expected to post high double-digit growth both sequentially and year over year in the third quarter, the company said. Foxconn also expects smart consumer electronics revenue to increase during the July-September quarter as the traditional peak season begins.

The company is preparing to increase AI rack production as NVIDIA's Vera Rubin systems move into mass production. Foxconn said Vera Rubin racks are scheduled to enter mass production in the third quarter and that full-year AI rack shipments are expected to more than double.

Foxconn had already demonstrated full-system NVIDIA Vera Rubin NVL72 racks at NVIDIA GTC in March, highlighting its manufacturing and system-integration capabilities. NVIDIA later said Vera Rubin was ramping into full production across more than 350 factories in 30 countries, with Foxconn among the global manufacturers involved.

The company is also expanding its networking business around the same AI infrastructure buildout. Revenue from 800G and higher-speed switches is expected to double for the full year, while co-packaged optics (CPO) switches are scheduled for mass production and shipment in the third quarter.

Foxconn's move into broader AI infrastructure also includes its collaboration with Schneider Electric to develop ready-to-deploy data center solutions combining rack integration, power systems, cooling and energy management.

Foxconn Plans More US AI Capacity

Foxconn expects demand for AI production capacity to remain strong in 2027 and plans to increase capital expenditure across Taiwan, the United States, Mexico and Vietnam.

In the United States, the company identified Texas, Wisconsin, Ohio and California for expanded research and development and manufacturing capabilities. The investment is tied to what Foxconn described as growing localization requirements in the US.

The expansion comes as other manufacturers are also adding US capacity for AI hardware. Wistron recently opened a $700 million manufacturing facility in Fort Worth, Texas, to produce components for advanced AI systems, including NVIDIA's GB300 and Vera Rubin platforms. 

Foxconn's first-half capital expenditure increased 5% year over year to NT$80.9 billion. At the same time, adjusted EBITDA rose 52% to NT$227.7 billion, which Chiang said was supporting the company's higher investment and financing costs.

Foxconn expects the third quarter to deliver significant sequential growth and strong year-over-year growth. Its full-year 2026 outlook remains unchanged, supported by AI server demand and growth in smart consumer electronics.

The company's results add another indication of how AI infrastructure spending is becoming a major growth driver for hardware manufacturers. Dell, for example, reported $16.1 billion in AI server revenue in its first quarter, exceeding revenue from its PC business.