Wiwynn, a prominent Nvidia server manufacturer, has raised alarms about the expanding AI infrastructure bottlenecks, stating that shortages of critical data center components are now prevalent beyond just memory chips. Chair Emily Hong noted that these supply constraints are likely to persist, with strong demand for AI hardware expected to last for the next 3 to 5 years. She anticipates that some alleviation in supplies may not occur until late 2027 or 2028. With over 80% of its sales coming from U.S. customers, Wiwynn is currently expanding its operations in Texas, which includes an already operational plant in El Paso and plans for additional facilities in the coming years.

Wiwynn: Wiwynn is a Taiwanese company that designs and manufactures servers and integrated data center infrastructure solutions focused on AI, cloud computing, and hyperscale deployments, with particular expertise in Nvidia-powered systems and advanced cooling technologies. It provides end-to-end hardware and cluster integration services to major technology clients worldwide. Chair Emily Hong recently addressed spreading component shortages and sustained AI hardware demand in the sector.
Emily Hong: Emily Hong serves as chairwoman of Wiwynn, guiding the company’s strategy in AI server manufacturing and global expansion. She has publicly discussed emerging bottlenecks in data center components and the long-term outlook for AI infrastructure needs based on direct industry engagement.

Supply Chain: Shortages of components such as high-end printed circuit boards, CPUs, multilayer ceramic capacitors, and power-related parts are affecting AI server production beyond memory chips.
Industry Trend: AI data center projects are encountering delays from limited supplies of electrical equipment including transformers and switchgear amid rapid buildout.
Manufacturing Strategy: Companies in the sector are prioritizing regional manufacturing resilience through expanded U.S. facilities to better serve hyperscale clients.