In this week’s techAsia briefing, we delve into the intensifying geopolitical pressures reshaping the global technology supply chain, from massive AI infrastructure investments to strategic manufacturing shifts.
The booming demand for artificial intelligence infrastructure continues to drive unprecedented capital expenditures across Asia’s top hardware manufacturers. Server giants like Quanta have significantly boosted their capital spending plans to expand capacity in California, Thailand, and Taiwan, anticipating doubled AI server output by year-end with orders already secured well into 2028. Similarly, power supply leader Lite-On Technology and chip substrate giant Unimicron have hiked spending to record levels, while Foxconn anticipates a 30% increase in capital expenditure.
This aggressive expansion has highlighted critical bottlenecks in the production equipment supply chain, with lead times for essential testing and weaving equipment stretching to between 30 and 50 weeks. Compounding these delays are China’s export controls on key raw materials and rare earths, alongside rising metal prices.
Against this backdrop of rapid expansion, companies are grappling with the uncertainties of U.S.-China geopolitical tensions. Executives face difficult strategic debates over whether to decouple their supply chains entirely from Chinese vendors to ensure continuity, raising profound questions about the necessity and feasibility of complete economic separation at such a granular level.
Pixel shift
Google has reportedly informed suppliers of its plans to halt production of all Pixel hardware—including smartphones, smartwatches, and wireless earbuds—in China next year. This strategic pivot follows the successful trial of higher-end Pixel production in Vietnam, positioning Google as the second major smartphone manufacturer, after Samsung, to establish a mature non-Chinese manufacturing ecosystem. Despite memory chip supply constraints, Google aims to increase Pixel shipments by 8% to 10% this year to capture greater market share.
Bit by bit
In a move to support its domestic AI leaders, Beijing has permitted small batches of Nvidia’s H200 processors to enter mainland China. Major players like ByteDance and Tencent have each received approximately 10,000 units recently. While the H200 represents at least two generations behind Nvidia’s most advanced, restricted chips, the U.S. licenses allow shipments to both mainland China and Hong Kong. However, Hong Kong’s limited data center capacity due to power supply constraints presents a significant logistical hurdle for local tech giants looking to utilize these advanced processors.
Open rivalry
The debate over open-weight AI models has emerged as a new flashpoint in U.S.-China technological rivalry. Following the release of China’s DeepSeek V4 Pro model, Washington is grappling with potential restrictions on Chinese open-weight AI. While critics warn of potential security backdoors or biases, many U.S. tech companies strongly oppose a ban, highlighting the utility of these freely available models for developers.
Material concerns
Beijing has intensified supply chain pressures on Taiwan by restricting or delaying exports of critical aerospace and optical materials, including germanium- and quartz-based components. Taiwanese manufacturers in the optical technology and semiconductor equipment sectors have faced prolonged customs delays and order cancellations. As China weaponizes its dominance in these critical raw materials, the geopolitical friction continues to cause tangible bottlenecks for Taiwan’s key high-tech industries.
Also Read
- VF Corporation Director Richard Carucci Acquires 20,000 Shares Valued at $295,000
- Taiwan Semiconductor vs. ASML: Choosing the Superior Semiconductor Investment
- Colombia and Israel Establish Mutual Visa-Free Travel as Diplomatic Ties Rebuilt
- S&P 500 Surges 33% Post-Election: Why Midterm Volatility Shouldn’t Derail Your Portfolio

