Thailand faces a critical window leading up to 2026 to secure its position as Southeast Asia’s sustainable data center hub, driven by a projected need for 3 to 5 gigawatts of capacity by the end of the decade.
The country’s current Direct Power Purchase Agreement (PPA) pilot program supplies approximately 2,000 megawatts, highlighting a gap that infrastructure providers must address to support growing artificial intelligence (AI) workloads. The transition to AI computing requires significantly higher power consumption than traditional computing models. Standard data center deployments typically operate at 4 to 5 kilowatts (kW) per rack, whereas ultra-high-density racks designed for AI can exceed 80kW.
According to Budsarin Pradityont, Country Head at ST Telemedia Global Data Centres (STT GDC) Thailand, providing clean and reliable energy at the speed required by AI expansion is the central challenge for the nation. She noted that the design choices made during a facility’s initial development establish its carbon emissions profile for 15 to 20 years of operation.
The push for sustainable digital infrastructure coincides with Thailand’s recent policy shift to move its Net Zero target forward to 2050, accelerating the previous goal of 2065 by fifteen years.
Corporate operators in the region are subsequently adapting their infrastructure to meet both technological demands and climate goals. STT GDC recently announced that its Bangkok 1 facility achieved certification under the NVIDIA DGX-Ready Data Center program, confirming its infrastructure is engineered to support advanced AI systems. The facility utilizes both liquid cooling and high-efficiency air-cooling to accommodate diverse AI architectures.
Regionally, STT GDC recently partnered with LITEON to launch Southeast Asia’s first High Voltage Direct Current (HVDC)-powered AI testbed in Singapore. The project demonstrated that HVDC architectures can yield up to 30% in energy savings and reduce power footprints by up to 40% in high-density environments.
By establishing an energy-efficient digital backbone, industry leaders aim to aid local enterprises and researchers in developing high-performance applications, such as the Thai-language large language model (ThaiLLM) initiative, while mitigating the environmental impact of expanding data volumes. This approach is intended to strengthen Thailand’s digital infrastructure while remaining aligned with global decarbonization targets.
51Talk marks 15 years, targets Thai market
Grab Thailand accelerates EV push with new partnerships
