(Sept 3): There is a global competition to develop artificial intelligence (AI). However, AI development goes beyond software and algorithms and is enabled by advanced hardware components, including semiconductor chips, hard drives, network devices, and other computing peripherals. All these are ultimately housed in data centers that make up cloud computing infrastructures.

Given the complex manufacturing processes and globalisd production networks that drive intense competition to build and secure hardware, a reassessment of the global supply chain strategy is necessary. Traditionally, technology companies have prioritised low-cost locations for sourcing or manufacturing hardware components. In light of AI’s strategic importance, acute material shortages, and vulnerability to disruptions, companies must align supply chain strategies with overall business value chain objectives, with emphasis on resilience, security, and sustainability over cost reduction alone.

The AI hardware value chain encompasses multiple stages: research, design, development, manufacturing, and distribution. Table 1 details the component groups within each stage. To prevent delivery hiccups at each stage of the value chain, the supply chain requires a reliable, secure flow of raw materials and resources from upstream to downstream. As such, there is a growing need for proximity to tighten the supply chain and cluster activities for better coordination of end-to-end operations. 

Table 1. Stages of the AI hardware value chain and their principal component groups

Source: Compiled by the authors from industry sources.

However, companies need to be cognisant that overconcentration of suppliers in co-location might create unintended risks, particularly energy, cyber, and physical security. For example, the recent Strait of Hormuz crisis caused an acute shortage not only of liquefied natural gas but also of helium, a critical gas in the photolithography process for integrated chip production, leading to a bottleneck in advanced chip production, especially for TSMC. In addition, service disruptions occurred across banking and other sectors after drones hit three of Amazon's data centre in the United Arab Emirates and Bahrain. In an unrelated case, Apple’s primary contract manufacturer, Foxconn, experienced a cyberattack in its North American facilities. These security issues amplify the impact of a single event and highlight the importance of a more carefully thought-out location-specific supply chain strategy.

Rather than focusing on a single location, an optimal sourcing strategy would be to limit geographical dispersion by nearshoring across countries within the same region. This regional spread is particularly important to ensure reliable and sustainable sourcing. It also offers alternative or multiple sourcing, enabling collaboration and competition whilst diversifying away from reliance on a single supplier and location.

Southeast Asia comprises 11 countries located between India and China. As a region, it is rising fast to become the fourth-largest economy globally, and a major driver of global economic growth. The region enjoys relative political stability and strong economic growth potential, and it promotes trade-openness policies, thereby mitigating geopolitical risks. Additionally, it is rapidly adopting digital technologies, with a young workforce and a relatively high literacy rate, essential for the labour-intensive, large-scale development of the manufacturing sector to progress along the value chain.

In light of the US-China trade conflict, which is driving the decoupling of the AI supply chain, Southeast Asia is emerging as a viable alternative or complementary source to a China-centric focus, offering a “China Plus One” strategy. Companies can maintain operations in China while concurrently expanding or relocating activities to Southeast Asia.

However, China Plus One is only the most immediate reason to look at Southeast Asia; the more durable reason is structural.

Strategically, Southeast Asia's diverse economic profile offers technology companies a critical opportunity to build a long-term competitive advantage. Historically, the region began developing its high-tech electronics manufacturing sector over half a century ago, driven by Singapore in the 1960s, Malaysia in the 1970s, and Thailand in the 1980s through foreign direct investment. The sector's maturity is due to strong multifaceted government support through trade policies, innovation funding, tax incentives, and other measures. Mature manufacturing operations are a critical advantage, lowering barriers to entry during sourcing shifts.

Singapore has developed a comprehensive semiconductor ecosystem with a comparatively strong focus on semiconductor research, chip design, and front-end wafer fabrication. On the other hand, Malaysia has an established electronics and semiconductor manufacturing base, and excels primarily in back-end assembly, test, and packaging. Thailand has developed into a global manufacturing hub for hard disk drives and the region's base for printed circuit boards. Meanwhile, Vietnam has benefitted significantly from the “China Plus One” strategy, owing to its proximity to China and the spillovers from the US-China trade conflict. For instance, OpenAI’s decision to produce its consumer hardware in Vietnam evidences its rapid evolution into a hub for manufacturing consumer electronics and computer components. Indonesia is leveraging its abundant natural resources, such as energy, water, and land, and aims to become the region’s data centre hub. The Philippines, meanwhile, is catching up with its version of AI Silicon Valley through the US-backed Pax Silica project. 

Table 2 summarises these country positions across the region.

Source: Compiled by the authors from industry sources.

Southeast Asia’s rise in strategic importance is not episodic. Its AI supply chain is demand-driven, catering not only to the global market but also to rapidly rising domestic consumption. Thus, it is more systemic and scalable than if it were supply-driven. According to McKinsey & Co, Southeast Asia's AI adoption momentum surpasses the global average. What drives this momentum is the growing embedding of AI by enterprises operating in Southeast Asia. It will then spur domestic innovation and investment in AI and cloud services, and in turn, generate demand for co-located data centres. Having AI’s hardware supply in strategic proximity can easily fulfill this demand. 

This is what gives the region a credible claim to becoming a strategic hub in the global AI supply chain.

The next article will dive deeper into Southeast Asia as a strategic hub for the global AI supply chain in the coming decade.

Dr Andrei OJ Kwok is associate professor at Monash University Malaysia and Dr Neale O’ Connor is associate professor at LaTrobe University, Australia.