Short-Term Volatility Inevitable, Impact Limited
Nvidia founder Jensen Huang's resignation has shocked the global AI industry, supply chain and markets. Acer Group founder Stan Shih said on the 13th that short-term industry volatility is inevitable, and Taiwanese manufacturers and the newly appointed management may need a period of adjustment, but Taiwan has irreplaceable core capabilities in chips and servers and is a trustworthy partner, so the overall impact on Taiwan will be limited, and the broader AI development trend will not change.
Taiwan Orders to Continue, Focus on Medium- to Long-Term Direction
Speaking at an event where the Stanshih Foundation signed a memorandum of cooperation with Kumamoto University in Japan, Stan Shih said the new appointees clearly understand they must rely on Taiwan's core capabilities, and Taiwan is currently irreplaceable in both chip and server production and supply. Global supply chains are inherently about division of labor and cooperation, and Taiwanese manufacturers have never sought winner-take-all, but uphold the philosophy of creating value through cooperation and shared evolution, making them the most trustworthy partners and difficult to replace.
Regarding downstream supply-chain production and orders, a top executive at a system maker said the key executives responsible for Nvidia's supply-chain management and procurement are Executive Vice President of Operations Debora Shoquist, and Executive Vice President and Chief Financial Officer Colette Kress, who is serving as interim CEO. In the short term, Nvidia's procurement from Taiwan-linked suppliers and capacity coordination are expected to continue following established project plans, cooperation items and orders, while the medium to long term will depend on changes in internal operating strategy and direction. Another senior hardware supplier executive said the change involves U.S. dominance over the global AI industry narrative, China's AI development progress, as well as growing opposition to AI infrastructure after the midterm elections and a more pessimistic outlook for data center buildout feasibility, and a widening production-supply gap will make cloud and AI computing resources increasingly expensive.
AI Trend Unchanged, Bubble Will Eventually Correct
On talk of an AI bubble, Stan Shih said the broader AI development trend will not change, only the pace of development must adjust to actual conditions, and if progress is too fast and investment returns disappoint, the market economy itself will keep correcting. Citing the development of the internet, he said excessive expectations create bubbles, but substantive development never stopped, and bubbles are inevitable in transitional periods while substantive development does not stop. He stressed that the key in the AI era is how to prevent AI from doing bad things and wrong things, which is why he seeks solutions through his Wangdao philosophy.
On short-term market volatility, Stan Shih said it was a normal correction in a market economy, and stock rises and falls are normal, but the current market is not rational. Citing rising memory prices as an example, he said the main cause was AI-driven demand growth outpacing capacity expansion, plus artificial speculation creating abnormal profits, which did not come from sufficient value creation and therefore cannot last. In the future, bottlenecks must be broken through with a brand-new problem-solving mindset, such as AMD's new architecture, 3D XDRAM, XRAM and other technologies, as semiconductors face a dynamic problem of insufficient memory supply, and if mainland Chinese technology catches up in the future, South Korea's monopoly will also be challenged.