The artificial-intelligence boom is creating demand far beyond software companies. It is reshaping the businesses that manufacture the servers, racks and components behind every major model.
Taiwanese electronics manufacturer Wistron launched an offering of 25 million global depositary shares worth as much as $1.48 billion, according to a term sheet reviewed by Reuters. Each depositary share represents 10 common shares in the company.
The proposed price range of $58.67 to $59.79 represented a discount of approximately 4% to 5.8% from Wistron’s September 7 closing price in Taiwan. The company intends to use the proceeds to purchase raw materials denominated in foreign currencies.
The structure allows Wistron to access international investors while raising U.S.-dollar funding for a supply chain that operates across currencies and continents.
The fundraising arrives during a rapid expansion period. Wistron is an important supplier of AI infrastructure, including systems connected to Nvidia’s latest hardware. It opened a manufacturing facility in Texas during July following an investment of approximately $700 million.
The company has also approved additional spending for facilities in Hsinchu and Kaohsiung, Taiwan, as well as its American subsidiaries and a Vietnam property business.
Wistron reported second-quarter revenue of NT$895.4 billion and profit after tax of NT$14.8 billion. Reuters calculated that quarterly revenue rose 64% from the previous year and net profit climbed 128%, driven by AI-server demand from cloud-computing providers and business customers.
Those results help explain why the company is willing to issue new equity despite the discount. When orders exceed manufacturing capacity, raising capital can accelerate procurement and expansion. The tradeoff is dilution: existing shareholders own a smaller percentage of the company after new shares are created.
Wistron’s share price had gained about 30% during 2026 before the latest offering, according to Reuters. The increase suggests investors have rewarded its exposure to AI infrastructure, although it does not guarantee that current growth rates will continue.
Hardware suppliers face significant risks. Manufacturing expansion is expensive, customers can delay orders, and generations of AI chips change quickly. A factory optimized for today’s systems may require additional investment when the next architecture arrives.
Still, Wistron’s fundraising shows where AI capital is flowing. The industry’s most visible products may be chatbots and software agents, but those tools depend on warehouses filled with specialized machines.
The race to build intelligence is also a race to finance metal, memory, cooling systems and manufacturing space.
