What progress have you made on Digital InfraCo’s strategic priorities set out in Singtel28?
Our target is to double Digital InfraCo’s EBIT by 2028. We have made decisive investments to capture growth opportunities from digitalisation, cloud and AI trends in Asia.
First, we are expanding Nxera’s footprint to meet the region’s rising demand for sustainable AI-ready data centres. By the end of 2026, our data centre capacity across Singapore, Malaysia, Thailand and Indonesia will more than double to over 200MW. As the data centres come online and are progressively turned on for customers in the next two to three years, we expect Nxera to nearly double its EBITDA by 2028. These next-generation facilities are built for AI cloud environments and are highly efficient in energy and water use. They will be hyper-connected through our regional telcos but carrier-neutral, giving customers flexibility and resiliency.
Second, we see strong potential in RE:AI, our AI cloud service. It enables organisations to deploy and scale GPU-powered AI applications without needing in-house expertise or costly infrastructure investments. This helps customers innovate faster and sharpen their competitiveness. By offering AI-as-a-Service (AIaaS), which integrates GPU computing power with AI tools and capabilities, and going beyond renting GPU time with just GPU-as-a-Service (GPUaaS), we can reach a wider range of industries and democratise AI access through RE:AI.
Third, our Paragon platform has the distinction of being the industry’s first multi-network, multi-cloud aggregation and orchestration platform for 5G and edge cloud computing. Designed to simplify the adoption of these technologies, Paragon is already used by organisations in sectors like manufacturing, healthcare and aviation. It also helps telcos monetise their application programming interfaces (APIs) and pursue private 5G network opportunities.
Our strategic investments in Nxera, RE:AI and Paragon have led to higher operating expenses in FY2025 as they are in the growth phase. From FY2026, we expect Nxera to deliver healthy growth, starting with our 25MW data centre in Thailand, which began operating in June 2025, and our 58MW DC Tuas in Singapore, expected to launch by January 2026.
Nxera was launched more than a year ago – how far has it come since?
The market’s response to Nxera’s hyper-connected, AI-ready and sustainable data centres has been very encouraging. Customers have already committed to a substantial amount of capacity in Thailand and Singapore ahead of the launch of our two data centres. We are also evaluating future phases based on customer demand and exploring entry into new markets such as Japan with our partner Hitachi.
There is strong demand in Asia for high power density data centres suitable for AI workloads, but supply remains limited. Our partnership with global investment firm KKR has helped us scale rapidly to address this gap and support the region’s digital and AI growth.
Nxera was built on our extensive experience in designing and running high-performance data centres in Singapore for hyperscalers and enterprises. Our flagship DC Tuas, which goes live in 2026, will be the most hyper-connected and energy-efficient data centre with the highest power density in Singapore, specifically built for intensive AI and cloud needs.
A key differentiator for Nxera is our commitment to sustainability. We aim to achieve operational net-zero emissions by 2028. We already use renewable energy in our operations and are securing more through partnerships to support customers in meeting their own green goals.
How much interest has RE:AI received, and how will you keep your edge?
RE:AI has received strong interest from enterprises and public sector agencies looking to become more efficient and intelligent through AI adoption. We launched this AI cloud service to address the barriers they face in adopting AI such as high costs, long GPU lead times, energy needs, data sovereignty concerns, the complexity of AI technology and a lack of in-house AI talent.
RE:AI is among the first AI cloud service providers in this region to offer NVIDIA’s new GB200 GPUs, delivering a leap in AI and high-performance computing. This sets us apart in a fast-evolving landscape. Our AIaaS offering is complemented by our GPUaaS offering, which went live separately last October and is attracting customers with established AI teams. Both AIaaS and GPUaaS are supported by our advanced network orchestration, which spans 5G, low latency, fixed and subsea cable networks, enabling customisation based on client needs. These advanced network capabilities and platform differentiation give us a clear edge in delivering latency-sensitive inference AI use cases.
What are the biggest opportunities and challenges shaping Digital InfraCo’s next phase of growth, and how will it stay ahead in a fast-evolving digital landscape?
We are focusing on deepening and expanding our data centre presence, strengthening our position in Southeast Asia while entering new markets such as Japan in Asia Pacific. At the same time, we must navigate increasingly complex regulations that could affect GPU allocations and influence data centre demand.
RE:AI is a key long-term investment for us. We have seen strong demand for sovereign AI cloud infrastructure in Singapore and plan to roll out similar offerings in other markets. This supports our ambition to become a leading regional AI enabler. Attracting and retaining skilled AI and data centre talent will be one of our toughest challenges as competition intensifies. This is also why we set up the Sustainable AI DC Academy and the AI Acceleration Academy with Institutes of Higher Learning to retrain our workforce and develop the talent pipeline to support our growing needs.
I am particularly proud that Paragon was recognised as the top all-rounder orchestration platform globally, outperforming more than 10 other network API exposure infrastructure vendors. This assessment underscores our unique position as an industry player with expertise in not only orchestrating cloud infrastructure and networks but in software applications as well.
We now hold six international patents for Paragon in the US, Singapore, Australia, South Korea, China and Indonesia. These patents protect our innovation in enabling the differentiation of 5G, edge computing and AI services through the platform. Paragon has evolved recently to empower telcos to monetise network APIs and establish sovereign AI factories within their respective countries.
Overall, we are optimistic about our growth outlook. Our investments in infrastructure and platforms have firmly established us as a regional leader. With our capabilities in place, we are well-positioned to capture the significant opportunities emerging from digitalisation, cloud and AI.
