T.Media logo

Custom AI Chips Drive Record Sales at Wiwynn

Reporter Richard Brown
Release time:2026/11/09 13:56
Last update time:2026/11/09 13:56
  • S

  • M

  • L

Wiwynn designs and integrates systems around both GPUs and custom ASICs.
Wiwynn designs and integrates systems around both GPUs and custom ASICs. (Image courtesy of Richard Brown.)

Cloud service providers are increasingly buying AI servers built around their own chips as well as Nvidia's, and Taiwan's server makers are taking orders for both. Wiwynn, a supplier of servers and data center systems to major cloud service providers, said its third quarter was driven by faster shipments of servers using custom processors, including Amazon's Trainium 3, along with steady demand for general-purpose servers.

Custom processors, or ASICs, are chips designed for specific workloads. They give cloud operators another way to handle large volumes of AI computing alongside Nvidia-based systems. Wiwynn designs and integrates systems around both GPUs and custom ASICs. Wistron, the electronics manufacturer that controls Wiwynn, also builds AI server systems and components as well as conventional servers and PCs.

Wiwynn's third-quarter revenue reached a record NT$434.62 billion (about US$13.3 billion), up 62.9% from a year earlier and about 56% from the second quarter. Net profit rose 51.6% to NT$23.36 billion (US$714 million). Wistron also posted record quarterly sales of about NT$1.15 trillion (US$35.2 billion), more than double a year earlier and up more than 27% from the previous quarter. Wistron includes Wiwynn's results in its own reported figures, so Wiwynn's revenue is counted again inside Wistron's total.

Demand held up after the quarter closed. Wistron's October revenue was NT$402.61 billion (US$12.3 billion), up 117.6% from a year earlier. The company expects AI-related growth to continue in the fourth quarter, with larger customer deliveries concentrated in November and December, while notebook shipments, already hurt this year by higher memory prices, face seasonal weakness.