Over the past six months, I’ve engaged with customers, partners, and governments across Asia, and one topic dominates every conversation: AI. Organisations are eager to leverage AI to enhance competitiveness, streamline operations, and accelerate innovation cycles. They view AI as a tool for extracting insights from data, improving decision-making, and addressing challenges such as scalability and workforce augmentation.
For governments across Asia-Pacific, the stakes are even higher. AI is seen as a cornerstone for economic growth by creating smarter cities, optimising public services, and bolstering digital economies. It promises infrastructure improvements, national security enhancements, and high-value jobs in emerging tech sectors.
A geopolitical imperative also drives the urgency. Nations aim to lead in AI research and development while ensuring control over critical technologies, data, and infrastructure. This pursuit of digital sovereignty safeguards independence and innovation capacity, cementing AI’s role as a mission-critical driver of strategic growth.
Yet, here’s the sobering reality: A significant portion of the world’s AI-processing chips resides outside Asia-Pacific, primarily in US data centres. For governments and enterprises, this degree of digital dependency is increasingly unsustainable as AI becomes a foundational element of business and national strategies.
In response, AI factories are emerging across Asia, from Jakarta to Johor Bahru. These specialised facilities aim to reduce reliance on US-based infrastructure and enable scalable, autonomous AI operations. They represent a shift toward strategic autonomy, offering localised innovation and adaptability to regional needs.
But as I see it, Asia’s rush to establish these advanced factories is uncovering deeper, systemic infrastructure challenges — challenges that are only now coming into sharper focus.
Beyond traditional data centres
Calling AI factories “data centres” oversimplifies their complexity. These hubs of computational power can consume energy equivalent to a small neighbourhood — or even an entire urban district at peak operation.
Meeting such energy demands is especially challenging in densely populated areas with limited power capacity. Regions like Johor Bahru and Jatiluhur are addressing this by harnessing renewable energy sources, such as solar, hydro, and geothermal power. This approach alleviates pressure on power grids, ensures scalability, and aligns with sustainability goals. Moreover, it enhances energy security by reducing reliance on imported fossil fuels, positioning these regions as pioneers in sustainable AI ecosystems.
Cooling these facilities is another challenge. Traditional air-cooling systems can’t cope with the heat generated by AI workloads. Liquid cooling systems, adapted from semiconductor facilities, are emerging as a game-changer, making data centre designs more efficient and future-proof. In tropical climates, this shift is essential for survival.
The new architecture of AI operations
AI factories orchestrate vast data flows between computational hubs, local AI models, and business applications. This hybrid architecture is shaped by Asia’s regulatory environments, diverse markets, and business priorities.
One transformative trend is data gravity, which is reshaping AI infrastructure. Unlike the cloud era, where data was sent to distant centralised resources, the AI era emphasises computational resources near the data itself. By localising capabilities like large language models, organisations ensure compliance with data sovereignty laws while unlocking real-time insights.
This proximity reduces latency and minimises the cost of data movement, enabling industries to innovate faster. Financial institutions process transactions with near-zero latency, healthcare providers analyse imaging datasets in real time, and manufacturers optimise quality control on production floors — all while maintaining centralised oversight.
Building the foundation
Creating seamless AI factories requires more than just advanced hardware. In my experience, their success hinges on addressing a critical skills gap. Overcoming infrastructure challenges demands rebuilding expertise in networking, storage, and hybrid systems to support these complex environments effectively. Over the past decade, many organisations deprioritised these skills, focusing on cloud management instead. However, the rise of hybrid architectures demands a revival of deep technical knowledge.
To bridge the gap, organisations are increasingly turning to sophisticated application delivery systems to integrate AI with business operations. These systems go beyond traffic management, ensuring reliability, security, and performance at scale. They intelligently route data, optimise resource use, and ensure compliance with regulatory frameworks.
From our experience across Asia-Pacific, successful organisations consistently prioritise three critical capabilities in their AI infrastructure:
- Intelligent traffic management:
Smooth data flow is vital for AI factories. Advanced traffic management dynamically allocates resources, ensuring performance and reliability across hybrid environments. These systems balance workloads, prioritise latency-sensitive tasks, and provide insights into data patterns to preempt bottlenecks.
- Advanced security controls:
AI systems face unique security challenges, from data leakage to malicious inputs. Robust controls, such as encryption, activity monitoring, and prompt sanitisation, are essential. Governance frameworks ensure ethical AI use, transparency, and accountability, protecting organisations in high-stakes environments.
- API management:
Seamless integration between AI models and business systems unlocks AI’s full potential. API management ensures secure communication, monitors performance, and supports complex workflows across cloud, edge, and on-premises environments. Adaptive API orchestration further enables dynamic interactions between systems.
The road ahead
Success in the AI era isn’t about the largest AI models or most processing power — it’s about orchestrating seamless data flows across complex environments. AI factories are more than infrastructure; they are the foundation for a self-sustaining digital economy.
Asia, with its diverse markets and talent, is uniquely positioned to lead this transformation. By prioritising resilient, scalable AI infrastructures, the region can drive innovation, foster collaboration, and empower local talent to address regional challenges.
