RE:AI, the sovereign AI cloud business of Singtel Digital InfraCo, is further empowering enterprises with the first Asia Pacific deployment of NVIDIA DGX GB300 to enhance enterprises’ ability to process larger, more complex AI workloads at scale. The initial capacity, offered through RE:AI’s GPU-as-a-Service, has been fully committed ahead of deployment, reflecting strong demand for advanced AI compute.
Through RE:AI, organisations can access DGX GB300 within a sovereign cloud without having to build and maintain the underlying AI infrastructure. The sovereign environment gives them greater control over where sensitive data and workloads are stored and processed, helping address data security, integrity and residency requirements. This is particularly important for organisations operating in regulated environments, where the use of AI must meet stringent governance and compliance obligations.

Bill Chang, Chief Executive Officer of Singtel’s Digital InfraCo, said: “AI is advancing rapidly and the cost and complexity of keeping infrastructure current can become a barrier to wider enterprise adoption of AI. Through our GPU-as-a-Service, powered by next-generation technologies such as NVIDIA DGX GB300, we enable customers to access the most advanced compute capabilities through a sovereign cloud, without having to continually refresh their own AI infrastructure while ensuring complete data residency. This reinforces Singtel Digital InfraCo’s role as a leading sovereign AI enabler, and advances Singapore’s development as a trusted, innovation-led AI hub.”
Demand for GB300 is strong, with initial capacity fully committed. In Singapore, RE:AI has partnered with HTX to upgrade its sovereign AI infrastructure NGINE (Next Generation Infrastructure) with DGX GB300 GPUs, which power embodied AI and multimodal model training at the Home Team Humanoid Robotics Centre (H2RC).
“Industry partnerships like what we have cultivated with Singtel RE:AI and NVIDIA make a huge difference not only to how effectively we can build AI capabilities for public safety but also drive Home Team-wide adoption as part of our HTxAI Movement,” said Ang Chee Wee, HTX Chief AI Officer and Assistant Chief Executive (Digital & Enterprise). “Being able to scale compute securely and flexibly is important for both development and adoption. That is why working with RE:AI to upgrade our sovereign AI infrastructure, NGINE, with DGX GB300s, was the most efficient route to ensure we have the headroom to grow as we push the envelope in areas like embodied AI, and beyond.”
“Worldwide telecommunications and digital infrastructure providers are building sovereign intelligence factories to service enterprises in a trusted and secure environment,” said Ronnie Vasishta, senior vice president of telecom at NVIDIA. “With NVIDIA, Singtel will give their enterprise customers access to a full stack AI infrastructure to enable them to refine reasoning models and build new agents through a locally operated cloud.”
The growing sophistication of AI is driving demand for greater computing power. DGX GB300 supports use cases such as AI agents that analyse large amounts of information faster and carry out complex, multi-step tasks. It also supports advanced robotic systems and scientific research by accelerating the analysis of large datasets.
The deployment of DGX GB300 to RE:AI complements Singtel‘s AI infrastructure stack which spans its Nxera’s AI-ready data centres, extensive connectivity and patented Paragon orchestration platform. NVIDIA Mission Control will help RE:AI monitor and operate its DGX GB300 infrastructure at scale. Together, these capabilities provide the integrated infrastructure enterprises need to deploy, manage and scale AI securely.
