85-year-old Lin Tsung-chi, founder of Taiwan's King Slide Works, has topped the island's rich list. The reason — an unexpected but logical triumph of the "periphery" of the AI industry. Since the start of the year, the company's market capitalization has soared by nearly 280%, and Lin's fortune, according to Forbes estimates, has reached $20.3 billion, allowing him to overtake Foxconn founder Terry Gou.
King Slide has historically been known as a manufacturer of furniture hinges and drawer slides. However, the key business today is rail mechanisms for server racks. These inconspicuous components are critically important: they hold heavy computing equipment and allow it to slide out for maintenance without disrupting the delicate cooling system. By my estimates, the company controls about 80% of the global market for high-performance server rails.
87% Margin: Higher Than Nvidia and TSMC
Demand for King Slide's products is growing in parallel with the exponential construction of data centers and the adoption of more powerful AI systems. The logic here is simple: when a server rack costs millions of dollars, customers are willing to pay for mechanical reliability rather than skimp on it. This bet on quality has led to striking financial results. In the latest quarter, King Slide's gross margin reached 87%, whereas a few years ago it was around 50%. For comparison: Nvidia's figure stands at 75%, and TSMC's at 68%. The company's executive vice president, JC Wang, attributes this profitability to two decades of engineering development in a narrow niche.
New Architectures — New Rails
I expect this growth is just the beginning. Cloud giants are actively transitioning to their own AI accelerators, increasing equipment density, and implementing liquid cooling. Each new chip differs in dimensions, heat dissipation, and connectors, requiring a complete redesign of rack structures and, accordingly, new rails. Notable examples include Google, which received an option to purchase up to 58.97 million shares of Marvell Technology, and Alibaba, which plans to raise $10.2 billion for AI infrastructure development.
The market is also expanding through the energy sector: the declared capacity of gas power plants to power data centers in the U.S. has grown from 97 GW to more than 189 GW in six months. Even Nvidia's generational shift — scaling the Vera Rubin platform to gigawatt-level capacities — works in King Slide's favor.
However, competition is intensifying. According to Daiwa Securities forecasts, the company's share of supplies for Nvidia systems could decline to 75% next year, as the chipmaker diversifies its supplier pool. This will give data center operators leverage to negotiate prices. Nevertheless, King Slide's fundamental dependence on global AI infrastructure investment remains direct and sustainable.
My comment: The King Slide story is a classic example of how the AI industry creates fortunes not only in "chips and software" but also in seemingly primitive mechanical components. It's a reminder to investors: in the AI era, not only technology leaders win, but also those who ensure their physical reliability.