DayOne backs NVIDIA's AI data centre design standard

DayOne will support the NVIDIA DSX POD reference design, linking the Singapore-based data centre operator to a standard for AI infrastructure deployment.

It will use its proprietary integrated delivery model alongside the design, which is intended to align compute, power and cooling in AI data centre projects. The approach is part of a wider strategy centred on standardised design, integrated delivery and modular infrastructure across multiple markets.

Demand for AI infrastructure has risen as deployments have grown larger and more complex. That has increased focus on how quickly operators can deliver power, cooling and physical space while maintaining common technical standards across sites.

DayOne uses a standardised method for data centre development that includes modular prefabrication of power, cooling and base-build systems. Moving some mechanical, electrical and plumbing work into factory settings reduces development timelines and supports consistency across large projects, it said.

The company said this model allows it to complete data centre projects in eight to 12 months in Southeast Asia. That is shorter than conventional build cycles and reflects growing pressure on operators to bring AI-ready capacity online more quickly.

Growing AI Demand

The announcement highlights how data centre groups are adjusting construction and delivery models as AI workloads place heavier demands on infrastructure. Compared with traditional data centre deployments, AI facilities typically require more tightly integrated systems for electricity supply, cooling, networking and computing hardware.

For operators and customers, this has increased the importance of repeatable designs and construction methods that can be rolled out across regions. DayOne said modular infrastructure would play a key role in delivering next-generation data centre capacity as these requirements intensify.

DayOne, established in 2022, operates across 10 markets in Asia Pacific and Europe. Its footprint includes Singapore; Johor and Greater Kuala Lumpur in Malaysia; Indonesia; Thailand; Tokyo and Kyushu in Japan; Hong Kong; Finland; and Spain.

The company develops and operates high-capacity campuses for hyperscalers and businesses. Its model centres on developing, constructing and operating data centres in multiple jurisdictions through a common delivery framework.

Jamie Khoo, Chief Executive Officer of DayOne, said the move was consistent with the company's existing methods for data centre construction.

"At DayOne, we believe that prefabrication and modularization are key to delivering infrastructure at the pace required by the AI era," said Jamie Khoo, Chief Executive Officer of DayOne. "Our teams have spent years developing integrated platforms, products, and modules that enable us to deliver large-scale campuses at speed and scale, with the consistency and quality our customers trust. Supporting the new NVIDIA DSX POD reference design is a natural extension of that work."

Industry Alignment

NVIDIA also presented the arrangement as a response to the need for closer coordination of infrastructure elements as demand for accelerated computing grows. The reference design is intended to help customers deploy AI systems more quickly by bringing compute and facility requirements into closer alignment.

"Meeting growing demand for accelerated computing requires integrated infrastructure solutions, and the NVIDIA DSX POD reference design helps align compute, power, and cooling for faster deployment," said Charlie Boyle, Vice President of DGX Systems at NVIDIA. "DayOne's support for this reference design, combined with its modular infrastructure and integrated delivery capabilities, can help customers bring capacity online sooner and shorten the time to first token."

The tie-up reflects a broader shift in the data centre market as operators seek more standardised ways to build facilities for AI workloads. As power availability, construction schedules and engineering integration become more critical, delivery models that can be repeated across markets are becoming a stronger point of competition.