Singapore has launched a data centre testbed with the National University of Singapore and Nanyang Technological University for researchers and industry partners to develop and test cooling technologies customised for tropical environments.
The Sustainable Tropical Data Centre Testbed (STDCT) sited at NUS College of Design and Engineering, covering 770 square metres, stands as a flexible, full-scale live facility for scientists to experiment and validate cooling ideas.
It will also serve as a "de-risking" platform for companies to optimise new technologies in a realistic, tropical setting.
Data centres are power-hungry facilities with a large appetite for electricity, especially in hot and humid climates such as Singapore.
On average, about 40 per cent of a DC’s energy consumption goes into powering its cooling and ventilation systems. Efficient cooling technologies are essential for reducing operation costs and lowering the environmental impact of DCs.
The new initiative aims to reduce cooling energy consumption by up to 40 per cent, cut water usage by 30 to 40 per cent, and reduce carbon dioxide emissions by about 40 per cent.
Funded by the National Research Foundation, the test bed has attracted more than S$30 million in investments from 20 industry partners like Facebook, Dell Technologies, Intel, Keppel Data Centres, and Schneider Electric among others aiding in construction, providing technical advice, equipment, funds and scholarships for researchers.
STDCT has five research projects, including identifying ways to raise the recommended temperature of data centres.
One of these projects includes research by the NUS team in pioneering the direct chip hybrid cooling system, which consists of a high-performance hybrid sink design with two modes of cooling.
A research team led by the NUS is designing a unique heat sink coupled with immersion cooling for enhanced cooling performance and another project aims to validate the potential of a novel cooling solution that uses a high-performance hygroscopic material to significantly improve cooling efficiency.
In tandem, scientists from NTU are leading two projects - one to establish the optimum temperature and humidity setpoints for air-cooling data centres in the tropics, and the other to develop a digital twin to enable real-time performance modelling and prediction.
STDCT said it will establish more strategic partnerships with industry standards organisations and leading corporations to provide learning opportunities for students and industry practitioners.
A whitepaper will also be developed to provide recommendations on optimum DC design and operations, and this is expected to be released in the fourth quarter of 2024.
