Foxconn Posts 35% Profit Surge as AI Servers Overtake iPhone Assembly for First Time

Foxconn Achieves Historic Profit Surge Powered by AI Infrastructure

Foxconn, officially known as Hon Hai Precision Industry, reported a forecast-beating second quarter net profit of NT$59.97 billion ($1.86 billion), marking a 35% increase from the previous year1

1

2

2

. This represents the company's highest ever profit for a second quarter, surpassing analyst estimates of NT$58.38 billion3

3

. Consolidated revenue hit NT$2.53 trillion, up 41% year-on-year and 19% from the previous quarter, with July revenue jumping 54.2% to a record NT$946.5bn1

1

. The profit surge signals a fundamental transformation in Foxconn's business model, as AI hardware sales now drive growth more than traditional consumer electronics.

AI Servers Overtake iPhone Assembly in Revenue Mix

Foxconn crossed a critical threshold this quarter as its cloud and networking division, which builds AI servers, accounted for 51% of revenue for the first time1

1

5

5

. Smart consumer electronics, the division containing iPhone assembly, fell to 29% of total revenue1

1

. This marks a dramatic shift from the first quarter when cloud and networking stood at 48%1

1

. The company has evolved beyond assembling low-margin iPhones to manufacturing AI servers for Nvidia, along with electric vehicles and robots2

2

. Rotating chief executive Michael Chiang framed this transformation as structural rather than cyclical, telling analysts that cloud investment "will become the most critical growth driver for Hon Hai over the next few years"1

1

.

Booming Cloud Investment Fuels AI Server Demand

Governments and tech giants are pouring massive capital into building data centers that train and run AI tools such as chatbots, image generators, and task-executing agents2

2

. This surge in AI infrastructure investment has turbocharged Foxconn's business, with the company expecting AI server rack shipments to grow by a high double-digit percentage sequentially in the third quarter and more than double this year5

5

. Chairman Young Liu stated that continuing infrastructure buildout is being driven by a broad range of customers, from AI model developers and cloud providers to governments and enterprises5

5

. Michael Chiang emphasized that AI infrastructure demand remains steady, unlike consumer electronics which lives and dies by seasonal cycles1

1

. While cloud service providers and AI labs are the biggest buyers currently, Chiang noted that government and corporate adoption remains in its infancy1

1

4

4

.

Global AI Server Production Expansion Across 24 Countries

Foxconn is ramping up global AI server production by expanding to hundreds of sites across 24 countries, nearly doubling its earlier footprint1

1

4

4

. New capacity is being added in California, Texas, Wisconsin, and Ohio, alongside key hubs in Taiwan, Mexico, and Vietnam1

1

4

4

. Through its flexible "local for local" strategy, Foxconn aims to meet localized client demand, secure supply chains, and insulate operations from geopolitical risks4

4

. The next platform test involves Nvidia's Vera Rubin racks, which enter mass-production preparation this quarter and ship in the fourth, with Chiang expecting them to become Foxconn's major product next year1

1

. Morgan Stanley analysts described 2026 as a particularly strong year for downstream rack assemblers like Foxconn4

4

.

Market Share Pressures and Commoditization Concerns

Despite record revenue, Foxconn faces competitive headwinds as cloud providers diversify their supply chain. Morgan Stanley expects the company's high-end rack market share to fall to 39% this year from 51% in 2025, attributing the shift to buyers seeking alternative suppliers rather than declining Foxconn output1

1

4

4

. Industry analysts have cautioned that as AI server designs standardize, hardware risks becoming commoditized, lowering barriers for rival assemblers1

1

4

4

. Michael Chiang rejected this assessment, arguing that standardization "highlights the advantages of suppliers with scaled delivery and complete vertical integration"1

1

4

4

. Pointing to Foxconn's one-stop assembly capabilities, extensive automation, and in-house component manufacturing, Chiang said the company expects its market share to increase on next-generation platforms4

4

.

Margin Pressures Persist Despite Revenue Growth

While headline profits impressed, margin dynamics reveal ongoing economic constraints. Gross margin stood at 6.12%, down 0.21 percentage points year-over-year1

1

. Operating margin improved to 3.75%, up 0.60 points, while net margin sat at 2.37%, slightly below last year1

1

. Assembling servers remains a business measured in single-digit gross margins, where volume drives profitability1

1

. Earnings per share reached NT$4.27, up from NT$3.19 a year earlier, with first-half profit hitting NT$109.89bn, up 27%1

1

. The company forecast "strong YoY growth" in the current quarter and "a strong full-year outlook"2

2

.

Supply Chain Bottlenecks and Market Skepticism

Michael Chiang identified CoWoS, the advanced packaging technology from TSMC that AI chips depend on, as the real constraint on 2027 growth rather than demand1

1

. The market expects TSMC's CoWoS capacity to grow by more than 50% next year, but how much translates into rack shipments "will depend on chip supply," Chiang noted1

1

. Foxconn does not control that supply, highlighting a critical dependency outside its influence. Despite strong results, investors remain cautious—Foxconn shares are up 17% this year, significantly lagging the broader Taiwan index's 57% gain1

1

. The stock hit an all-time high in June before losing ground5

5

. Market skepticism centers on whether tech giants can sustain the AI infrastructure buildout, as Alphabet, Meta, Microsoft, and Amazon have committed close to $2.4 trillion to AI over coming years, with some financing through debt1

1

.