Power, grid limits test Vietnam's AI ambitions

As Vietnam aims to become one of Southeast Asia's leading AI hubs by 2030, experts warn that power shortages and limited grid capacity could constrain its ambitions.

"Electricity supply could become one of the most important constraints on Vietnam's AI ambitions," said Tran Minh Ngoc, a professor of business analytics at the University of Sydney Business School.

AI and data centers are highly electricity-intensive and require large amounts of reliable, uninterrupted power, Ngoc said. The challenge is particularly critical as Vietnam's AI strategy calls for major expansion of computing infrastructure, including about 500 megawatts of AI computing capacity by 2030.

Vietnam has seen a wave of data center investment in recent years as companies seek alternatives in Southeast Asia.

The country approved a national strategy aimed at making it among the AI leaders in the Association of Southeast Asian Nations by 2030, with the AI economy expected to account for 6 percent of GDP, Viet Nam News reported last month.

However, the Ministry of Industry and Trade said Vietnam could face a shortfall of about 4,256 MW of capacity and 2.9 billion kilowatt-hours of electricity as early as next year, as rising demand from data centers, AI and semiconductor manufacturing pushes up consumption and some planned power projects are delayed.

The shortfall is projected to widen to 20.7 billion kWh in 2028, 50.4 billion kWh in 2029 and about 61 billion kWh in 2030.

Although Vietnam has a strong foundation for developing large-scale AI data centers, doing so will require proactive planning for power systems and infrastructure, said Nguyen Duc Tuyen, head of the Power Grid and Renewable Energy Laboratory at Hanoi University of Science and Technology.

"Data centers are different from many conventional loads," said Tuyen, an associate professor and senior lecturer at the university's School of Electrical and Electronic Engineering.

"They can create very large, concentrated and relatively continuous electricity demand, and they require a very high level of reliability and power quality."

The government's review of the country's Power Development Plan VIII to reassess whether the planned capacity will meet future demand is timely, he said.

Vietnam should also move toward a "data center power supply road map" to develop a forward-looking strategy in line with the power system's capability and long-term energy transition goals, he added.

David Jackson, principal and CEO for Canadian real estate firm Avison Young in Vietnam and Cambodia, said Vietnam's data center boom is about more than real estate, as investors are seeking not only land but also places where the technology can operate at scale with reliable power and connectivity.

"The investment is coming, which is encouraging, but the bigger question is how much of the capability, knowledge and value stays here in Vietnam," Jackson said, adding that the projected 61 billion kWh shortfall can also be seen as a major infrastructure opportunity.

For example, he said a direct renewable power arrangement is being explored for a major new data center in Ho Chi Minh City, which also launched one of Southeast Asia's largest national data centers last year. The government has opened the direct power purchase mechanism to data center users, he added.

Reliability, affordability

James Kang, a senior lecturer in computing technologies at RMIT University Vietnam, said the future of power and grid support for AI data centers lies in reliability, additional capacity, affordability and efficiency.

The key question is not only how to secure enough electricity, but how to build an energy and digital infrastructure system that supports AI without raising costs or reducing availability for the wider population, he said.

Ngoc of the University of Sydney Business School also proposed making "Green AI" a strategic research and development priority, focusing on AI models, algorithms and computing systems that achieve strong performance with substantially less computing power and energy.

"Developing strength in Green AI could help Vietnam pursue its AI ambitions more sustainably while also creating a distinctive area of technological expertise," he said, adding that Vietnam could develop its expertise rather than compete only by building large facilities.

Contact the writers at kelly@chinadailyapac.com