South Korean circuit board maker Isu Petasys, whose clients include Alphabet, Nvidia, and Microsoft, has seen a 475%+ stock surge in 2023 fueled by the AI boom
Yoojung Lee / Bloomberg : Twitter: @uradn Twitter: @uradn : Isu Petasys Co., which counts Alphabet Inc., Nvidia Corp. and Microsoft Corp. as customers of its multilayered boards, has seen its share price surge 487% this year - among the biggest gainers on South Korea's benchmark stock index. https://www.bloomberg.com/...
Context & Ripple Effects
Isu Petasys shows how AI-server demand reached beyond the chip designers and cloud platforms to a South Korean supplier of multilayer circuit boards. Its customer list ties its valuation directly to infrastructure spending by Alphabet, Nvidia, and Microsoft.
The move preceded a broader pattern: SK Hynix’s AI-memory-driven share gains and later growth in South Korean chip output tied to AI-related memory demand similarly showed AI demand propagating through the country’s hardware supply chain.
First-order effects
- Isu Petasys gains a sharply higher market valuation as investors reprice its exposure to AI-server customers and multilayer-board demand.
- Alphabet, Nvidia, and Microsoft’s infrastructure build-out becomes economically meaningful for a component supplier rather than only for the companies buying and designing AI systems.
Second-order effects
- Other board and server-component suppliers face stronger investor scrutiny over their AI-server exposure, capacity, and customer concentration.
- The rally reinforces the read-through from AI compute spending to adjacent hardware categories, alongside the AI-driven rise in server maker Supermicro’s shares.
Third-order effects
- If AI-server deployment remains broad-based, more value may accrue to specialized upstream components whose performance and supply are integral to system assembly—not solely to leading chip vendors.
- The pattern points to a more geographically distributed AI hardware value chain, with South Korean suppliers participating through boards and memory as well as finished chips.
The trend: This is one data point in the AI infrastructure supply spillover, where spending by major AI platforms is transmitted through increasingly specialized hardware suppliers.