Taipei: The rapid growth of AI is transforming the role of server manufacturers, forcing them to expand globally and take on increasingly complex integration work that extends far beyond traditional hardware assembly, according to Wistron Corp. At a Taiwan Stock Exchange (TWSE) media briefing held alongside the 2026 Computex expo in Taipei on Thursday, Wistron CTO David Shen said AI infrastructure projects are becoming increasingly complex, requiring manufacturers to integrate computing, networking, storage, cooling, and semiconductor technologies.
According to Focus Taiwan, Shen highlighted that the scale and growth of AI infrastructure are remarkable. Wistron's revenue soared from NT$1.049 trillion (US$33.26 billion) in 2024 to NT$2.186 trillion in 2025. AI infrastructure's contribution to revenue escalated from 46 percent in 2024 to 71 percent in 2025, and further to 79 percent in the first quarter of 2026, as per company data. Shen attributed this rise almost entirely to AI-related business, with traditional PC operations showing more modest growth.
Shen also addressed the notion that Wistron's overseas expansion is mainly driven by tariffs. He clarified that the company is broadening its manufacturing, engineering, and research and development operations closer to customers in key markets, particularly in the United States. Recent facility establishments in Dallas, Fremont, and San Jose, along with expansions in Europe and Asia, underscore this strategy.
One of the most significant changes brought about by AI is what Shen described as "super integration." Unlike traditional server manufacturing, AI infrastructure demands original equipment manufacturers to coordinate technologies spanning networking, storage, cooling systems, power distribution, and advanced semiconductor packaging. As AI clusters grow larger and more powerful, manufacturers must collaborate closely with suppliers throughout the supply chain, including semiconductor companies, to optimize system performance, thermal management, and power efficiency, Shen explained.
Other companies participating in the briefing echoed the growing complexity of AI infrastructure. Cooling-system supplier Asia Vital Components (AVC) noted the accelerating demand for liquid-cooling technologies as potent AI chips generate more heat. Delta Electronics highlighted electricity supply as a growing constraint on future AI deployment and mentioned its development of solutions like microgrids, energy storage systems, and next-generation power architecture to help data centers manage rising energy demands.
The event was hosted by the TWSE as part of its efforts to showcase Taiwan's AI supply chain during Computex.