Juwai IQI Co-Founder and Group CEO: Malaysia Poised for Data Center Boom as DeepSeek Lowers AI Cost

Initial concerns over the sustainability of Malaysia’s data center expansion following DeepSeek’s emergence are now turning into optimism, according to an analysis by Juwai IQI Co-Founder and Group CEO Kashif Ansari.

Ansari noted that DeepSeek had disrupted the AI landscape with a large language model that rivaled those of established competitors like OpenAI at a significantly lower cost. He suggested that the affordability of AI models such as DeepSeek’s would likely drive even greater demand for data centers in Malaysia. Lower AI costs, he explained, would enable widespread adoption of AI-powered tools, which many businesses had previously avoided due to financial constraints. A survey conducted in November indicated that cost concerns remained the primary obstacle preventing businesses from integrating AI.

He further pointed out that large language models needed to become more affordable before AI could achieve mass adoption, adding that DeepSeek appeared to have made substantial progress in this regard. He cited the fact that its token pricing was just one-fifteenth that of its more established competitor, Anthropic. Given this drastic reduction in cost, he anticipated a surge in AI adoption, emphasizing that when AI services become cheaper, people tend to use them more frequently. This increased usage, he predicted, would create unprecedented demand for data centers, including those in Malaysia.

Supporting this outlook, the report referenced Morgan Stanley’s projection that power demand in the Asia-Pacific region, excluding China, could rise by 4% if technology companies accelerate investments due to reductions in AI’s energy consumption. For Malaysia’s real estate sector, this suggested that demand for land suitable for data center development would remain strong and might even increase.

However, not all data centers are equipped to handle the growing demands of AI. Many existing facilities, originally designed for enterprise IT, cloud computing, and general-purpose workloads, lack the power, cooling, and networking capabilities required for AI operations. Despite this, Malaysia is well-positioned to lead in AI-driven data center development, with billions of dollars worth of planned AI-ready facilities. Government initiatives and significant investments from global tech giants—including Nvidia, ByteDance, Google, AWS, and Microsoft—have helped shape this robust pipeline.

Malaysia’s data center development pipeline currently stands at 1.2 GW, marking a projected 600% growth over the next five years. Around 55% of upcoming projects are set for Kuala Lumpur, with the remaining 45% located in Johor. The report also cited Blackstone, one of the world’s largest AI infrastructure investors, which estimated that $2 trillion would be required for global data center expansion over the next five years. Malaysia, particularly Johor, was identified as one of the fastest-growing markets in Southeast Asia and well-positioned to capture a significant share of this growth.

The shift towards AI-driven data centers was already evident, with companies such as YTL Power incorporating advanced cooling technologies into their facilities. Its green data center park in Johor, for example, would feature direct-to-chip liquid cooling to support the high-power densities essential for AI processing.

Ansari emphasized that Malaysia remained an attractive destination for data center investments regardless of whether DeepSeek permanently reshaped the AI industry. He highlighted the country’s competitive advantages, including affordable land and energy costs, a skilled workforce, strong digital infrastructure, government incentives, and strategic connectivity to key markets such as Singapore, ASEAN, and China.

Energy costs, he pointed out, played a crucial role in operational expenses. He noted that Malaysia’s industrial power tariffs stood at approximately $0.10 per kilowatt-hour (kWh), significantly lower than Singapore’s rate of around $0.27 per kWh.

Ansari also suggested that Malaysia stood to benefit from AI’s increasing affordability in multiple ways. Previously, he said, the assumption had been that Malaysia would require billions of dollars to develop its own language model. However, DeepSeek’s claim of having created a similar model for under $10 million suggested that AI development was now well within Malaysia’s financial reach. At this price point, the country could afford to create its own language models and integrate AI into key sectors such as research, education, and the broader economy.

Beyond businesses, he added, Malaysian consumers would also gain from AI’s growing accessibility. Cheaper AI, he argued, would be embedded into everyday tools and services, making it more impactful in daily life. He compared AI adoption to the proliferation of mobile phones, noting that widespread access had only become possible once costs had declined significantly. Similarly, he envisioned a future in which AI-powered services would become as ubiquitous and transformative as electricity itself.

Have you read?
Countries: Powerful Passports.
Countries: Richest.
Countries: Poorest.
Countries: Happiest.
Countries: Life Expectancy.