Telekom Malaysia recently launched a GPUaaS platform to allow customers to tailor their infrastructure by selecting combinations of hardware, GPUs, and connectivity
In sum – what you need to know:
AI demand reshapes infrastructure – Cloud growth, immersive apps, and national AI initiatives are driving urgent demand for scalable, low-latency AI infrastructure across Southeast Asia.
TM lowers AI barriers – Telekom Malaysia’s GPU-as-a-Service tackles key adoption challenges like high upfront costs, data sovereignty, and inflexible infrastructure by offering localised, modular GPU access.
Malaysia rises as AI hub – With hyperscaler-ready data centers, a national fiber backbone, and regional connectivity, TM is positioning Malaysia as a launchpad for AI innovation across ASEAN.
Telekom Malaysia (TM), through its wholesale division TM Global, is positioning itself at the heart of Southeast Asia’s digital transformation by building out AI-ready infrastructure that meets the region’s evolving needs for scalability, data sovereignty, and energy efficiency.
As the region accelerates its adoption of AI technologies, the demand for low-latency, high-performance compute infrastructure has surged. In an interview with RCR Wireless News, Baharul Nizam Said Daliman, vice president of international sales at TM Global, outlined how TM is responding to this shift.
“The surge in cloud adoption and the expansion of digital services—from e-commerce to fintech and smart farming—require robust, scalable computing power to handle complex AI workloads,” Baharul Nizam said. He also pointed to a wave of government-backed AI initiatives, such as Malaysia’s National AI Roadmap and the National AI Office, which are helping to foster a strong AI ecosystem.
Solving real pain points with GPUaaS
TM’s GPU-as-a-Service (GPUaaS) platform, launched in ASEAN in December, is a cornerstone of this strategy. It was designed to tackle some of the most common barriers to AI adoption: cost, latency, and compliance.
“One of the most pressing barriers was the high upfront cost of AI infrastructure. Our service removes that barrier by offering scalable, on-demand access to powerful Nvidia GPU resources,” Nizam said. These resources are hosted in TM’s Uptime Tier III and LEED-certified data centers.
In addition to lowering the cost barrier, TM localised its GPUaaS platform to ensure data sovereignty and minimise latency. He added: “By localizing GPUaaS within Malaysia and integrating it with TM’s nationwide fibre network and strategically placed edge facilities, we deliver low-latency performance and ensure secure, in-country data residency—an essential requirement for regulated industries.”
The platform is modular and flexible, allowing customers to tailor their infrastructure by selecting combinations of hardware, GPUs, and connectivity. This approach supports a wide range of AI use cases — from diagnostic imaging and public sector avatars, to immersive content development and autonomous navigation.
Scaling for hyperscalers and enterprises
As hyperscalers expand their footprint in the region, TM is also upgrading its infrastructure to meet their growing needs. Its Nxera data center in Johor, set to launch soon, will offer up to 200 MW of scalable power and is designed to support energy-intensive AI and cloud workloads.
“Malaysia’s strategic location gives TM a unique advantage in supporting hyperscalers with a strong foundation of secure, high-performance infrastructure,” said Nizam. TM’s network includes a mix of core and edge data centres linked by robust fibre connectivity and backed by terrestrial and undersea cables across ASEAN.
For regional enterprises, scaling AI workloads locally presents its own challenges, such as limited access to high-performance compute and data governance concerns. TM aims to close these gaps with flexible infrastructure, including 29 global Points of Presence and over 160 CDN nodes, and a Facilities-Based Operator (FBO) license in Singapore.
Malaysia’s role in the regional AI future
Looking ahead, TM sees itself and Malaysia playing a central role in shaping Southeast Asia’s digital future.
“In the next 3 to 5 years, TM Global… will continue building on this legacy by playing a central role in enabling the growth of regional data center and AI infrastructure,” Nizam said. By combining infrastructure, connectivity, and governance, TM aims to create a trusted platform for cloud growth, AI innovation, and seamless cross-border data flows, the executive added.
TM Global is the wholesale division of Telekom Malaysia, handling both domestic and international markets. TM Global provides data center services including hosting and edge facilities. It also offers voice services, including bilateral and VoIP, as well as multimedia solutions.
