As Singapore plans for a sharp rise in electricity demand, Delta Electronics has formalised its closer collaboration with College of Design and Engineering (CDE) at National University of Singapore (NUS) to test more energy-efficient data centre technologies.

Announcing this at Delta SEAO Future Industry Summit 2026 in Pan Pacific Orchard, the company is set to take part in the Sustainable Tropical Data Centre Testbed Phase 2.0 (STDCT 2.0) Consortium.

It will provide NUS CDE with an AI Containerised Data Center and related equipment to support research and testing of data centre technologies suited to tropical conditions.

The collaboration combines Delta’s expertise in power, cooling and data centre infrastructure with NUS CDE’s research capabilities in tropical data centre design and testing.

Why energy is the issue

The announcement comes four days after the Energy Market Authority (EMA) unveiled a five-year Energy Infrastructure Plan to meet Singapore’s growing electricity needs.

Peak demand stood at 8.6GW in August 2026 and could reach 9.9GW to 13GW by 2031, which EMA attributes mainly to energy-intensive sectors such as semiconductor manufacturing and data centres, as well as wider use of electric transport.

Meeting that demand requires a major build-out. Five new gas-fired power plants will come online between 2026 and 2031, lifting generation capacity by around 26 percent from about 12GW.

Singapore is also investing up to S$5.3 billion between 2025 and 2029 in its power grid, 11 times the preceding five years. Natural gas supplies around 95 percent of Singapore’s electricity today, so how efficiently large new loads such as AI data centres use power bears directly on how much generation and grid capacity must be built.

That is where the testbed fits. AI workloads pack far more computing into each rack, which raises both power draw and heat. Cooling that heat in Singapore’s hot, humid climate adds to the load. Hosted at NUS CDE, STDCT 2.0 is Singapore’s platform for developing and testing tropical data centre designs.

“AI is fundamentally changing the design envelope for data centers, with much higher power densities and thermal loads placing new demands on power, cooling and infrastructure efficiency,” said Lee Poh Seng, Programme Director of STDCT 2.0 and Head of the Department of Mechanical Engineering at NUS CDE.

“Our goal is to generate solutions and evidence that can accelerate the deployment of more sustainable AI infrastructure in Singapore and across the wider region,” he added.

The question of power supply also featured in the Summit’s executive address Sustainability in the Age of AI and Energy Transformation.

“So this raises a critical question: Will power supply become the bottleneck for AI?” said Shan Shan Guo, Chief Brand Officer of Delta Electronics. “It is about building resilient, reliable, and efficient grid system, and this is exactly where microgrid can make a real difference.”

She highlighted how Delta has been developing microgrid systems for a decade, combining on-site generation, energy storage and smart management to provide more reliable power and reduce pressure on the grid.

Regional challenge

Singapore is not alone as data centre power demand is climbing across Southeast Asia. An IDC study commissioned by Delta found that operators in the region reported 26-50 percent year-on-year growth in power demand, against 11-20 percent in the US and Europe. About 60 percent are deploying, planning or evaluating direct liquid cooling, the highest rate of the three regions surveyed. Between 24 percent and 26 percent cite power, grid, cooling capacity and equipment lead-time risks as their single biggest constraint on scaling.

“Southeast Asia’s infrastructure needs are evolving rapidly as AI and electrification reshape the region,” said Jason Yuan, President of Delta Electronics Southeast Asia & Oceania.

Speaking in his session, Building Regional Readiness to Capture the Next Wave of Growth, on whether the region is ready to support the growth in AI infrastructure, he said the answer depends on three key factors: Power availability, energy efficiency, and speed of deployment.

Yuan closed by sharing the four stages for new capacity: plan and approve, design, build and commission, and operate and optimise. He pointed out that government, the private sector, utilities, and communities should be involved at every stage to deliver reliable, resource-efficient AI capacity.

Building on the December 2025 MOU

The move builds on a Memorandum of Understanding signed in December 2025 between Delta Electronics (Singapore) and NUS CDE.

The three-year agreement covers smart energy systems integration, advanced thermal management and modular infrastructure for challenging tropical environments.

The partners are also exploring Delta’s autonomous inspection robots for data centres to enhance safety, real-time monitoring and predictive maintenance. This echoes Delta’s support for Singapore’s Green Plan 2030 and Smart Nation goals.