Perspectives

Southeast Asia’s data centres and AI infrastructure imperative

Capitalising on a once-in-a-generation opportunity

Southeast Asia’s national governments, national players, global technology companies, and investors must act now to secure their ability to foster innovation, maintain economic competitiveness, and safeguard national security.

Artificial intelligence (AI) is amongst the most defining megatrends of this century, permeating our daily lives and driving substantial growth across economies. But not all AI is equal. Of note is generative AI (GenAI), which goes beyond mere pattern recognition capabilities of traditional AI to transform how images are generated, processed, and stored.

GenAI applications with image recognition capabilities, for example, are expected to replace sensors in many use cases, with significant and widespread repercussions for automation, mobility, and manufacturing, amongst others. These, in turn, must be supported by highly specific and demanding specifications for data centres and other AI infrastructure – most of which cannot currently be met in Southeast Asia today.

The emerging GenAI value chain

Along with the rise of GenAI is a value chain that is emerging to support it (see figure below). On the surface, it looks similar to the traditional AI value chain – with the exception of foundational models. But it is this very difference that drives differential value creation across the value chain.

Given the rapid pace of GenAI’s evolution, all players must continuously reassess their value creation models while adapting and responding to disruptions that can occur at any time. To this end, we believe that Southeast Asian players should consider their value creation activities across all three segments of the GenAI value chain: Application, Platform, and Infrastructure.

The emerging GenAI value chain

Move now, and move quickly


National governments, national players, global technology companies, and investors must recognise data centres and other AI infrastructure as critical assets of tomorrow – and move now, and move quickly, to build these assets on their shores.

National governments

National governments have arguably the most pivotal and important role to play in enabling the development of these assets and spurring growth in local and regional ecosystems by:

• Regulating and incentivising investments, by ensuring clarity in policies, standards, data security classification, and data localisation regulations, as well as providing infrastructure grants and incentives

• Attracting global players, including hyperscalers and cloud providers, by streamlining approvals, achieving policy stability, and ensuring infrastructure readiness

• Developing local and regional ecosystems with investors, local talent pool, and infrastructure players, including sovereign data ecosystems based on population size; economies with smaller populations should consider joining other ecosystems with similar profiles

• Managing risks, such as technology availability (particularly for GPUs in an export-controlled system), commercial (through close collaborations with national players on demand forecasts), consumer safety, data privacy, and cyber risks (e.g., Singapore’s Digital Infrastructure Act)

National players

National players should take concerted steps to consider how they can best support the build-out of data centres and other AI infrastructure on their shores by:

• Understanding where to play, including size of opportunity, which part of the value chain to target, and how much investment to commit

• Developing go-to-market strategies, to realise the market opportunity (e.g., through partnerships/alliances to monetise investments in infrastructure)

• Considering funding or co-funding models, including whether to go it alone or find a co-investor, and structuring the asset to unlock value

• Managing commercial risks, by developing accurate demand forecasts

Investors

As they consider how best to deploy their capital, investors and private equity players should think about:

• Understanding where to play, including size of opportunity, which part of the value chain to target, and how much investment to commit

• Weighing their partnership or consortia options to invest in data centres and other AI infrastructure

• Assessing new methods to reduce financing costs (e.g., real estate investment trusts (REIT) management and partial investment)

• Managing commercial risks, by considering a modular or phased build approach to reduce upfront costs and time-to-revenue

Did you find this useful?