Hanoi (VNA) – A wave of investment in artificial intelligence (AI) data centres is sweeping across Vietnam, with both domestic and international corporations announcing multi-billion-USD projects that are set to transform the country into a rising hub in the regional AI infrastructure map.
Recently, the Kinh Bac City Development Holding Corporation (KBC) signed a partnership agreement with Accelerated Infrastructure Capital (AIC) and VietinBank to develop a large-scale AI-integrated data centre at Tan Phu Trung Industrial Park in Ho Chi Minh City. The 10-hectare project will have a total investment of nearly 2 billion USD, featuring an AI campus with a designed IT load of 200 MW and the capacity to operate about 100,000 GPUs.
Earlier, HCM City received a proposal to develop a hyperscale data centre worth around 2 billion USD from a group of investors, including G42 (UAE), Microsoft (US), FPT, VinaCapital, and Viet Thai. The project aims to become an “AI Factory” providing advanced AI solutions and cloud infrastructure for Asia and beyond.
Viettel is taking the lead with two major AI data centre projects. In April, the group will begin construction of a 140-MW data centre in Tan Phu Trung, one of the Top 10 largest in Southeast Asia. The first phase will start operations in early next year, and the project is expected to be completed by 2030. In August, Viettel inaugurated the An Khanh Data Centre valued at 17.5 trillion VND (665 million USD), meeting Uptime Tier III standards and integrating Viettel’s in-house AI solutions.
The CMC Corporation is expanding its international-standard data centre chain with facilities, such as CMC Creative Space in Tan Thuan in HCM City, CMC Duy Tan Data Centre in Hanoi, and the upcoming CMC Hyperscale Data Centre in the city’s high-tech park, with a total investment of 250 million USD and an initial capacity of 30 MW, expandable to 120 MW.
A joint venture between VNG and ST Telemedia Global Data Centres is also building two AI data centres - STT VNG Ho Chi Minh City 1 and 2— scheduled to become operational in the first half of 2026.
According to Deloitte, Vietnam’s AI market could reach 65 billion USD by 2035, including 25 billion USD from AI data centre infrastructure. A report of the Vietnam Cloud Computing and Data Centre Club (VNCDC) shows that the domestic cloud market could hit 1.24 billion USD by 2025 and 2.5 billion USD by 2029, while total data centre capacity is expected to nearly double from 525 MW in 2025 to around 1,000 MW by 2030.
Vladimir Kangin, CEO of IPTP Networks, which is investing 200 million USD in an AI data centre in Da Nang Hi-Tech Park, said Vietnam offers huge potential. Early investors will seize the best opportunities while others who are waiting another five to ten years may be too late.
He noted that Vietnam currently has the region’s lowest data center construction cost per MW and profit margins second only to Singapore.
He recommended that the Government allow telecom operators to use existing power transmission lines for data cables, cut investment costs, and improve energy policies to encourage new power sources, including nuclear power, to support AI data centres.
Dang Tung Son, Vice Chairman and Chief Strategy Officer of CMC, added that Vietnam’s investment and operating costs for data centres are among the lowest in the region, about 6–7 million USD per MW, 40–60% lower than Singapore. Industrial electricity prices average 6–10 US cents per kWh, roughly one-third of Singapore’s level, making Vietnam a highly efficient investment destination.
In addition, Vietnam has an improving legal framework, a strategic location, robust connectivity, and growing data demand driven by AI, cloud computing, fintech, and e-commerce, which create a stable and promising market.
However, to further attract investors, Vietnam needs to ensure stable power supply, simplify project approval procedures, expand international transmission capacity, and develop high-quality human resources.
Dong Mai Lam, General Director of Schneider Electric Vietnam and Cambodia, said AI is transforming the design and operation of data centre. By 2028, AI could account for 15–20% of total data centre energy consumption, up from 8% in 2023, requiring more efficient cooling systems and higher computing density. Data processing is also shifting from centralised to distributed models, with about 50% of AI workloads expected to run on hybrid systems that combine core data centers with edge computing to boost speed and reduce latency.
To keep pace with the AI revolution, Vietnam needs a design, construction, and operation revolution for its data centres, Lam emphasised./.
