Southeast Asia is emerging as a key hub for artificial intelligence (AI) infrastructure as telecommunications providers, data centers, and emerging neoscalers invest in high-capacity connectivity to support the next phase of AI adoption.
Insights from a Ciena virtual roundtable with Francisco Sant’Anna, Senior Advisor for Marketing Intelligence at Ciena; Alex Wong, Regional Managing Director for ASEAN, Hong Kong, and Taiwan at Ciena; and Baharul Nizam Said Daliman, Vice President for International Sales at TM Global, highlighted how AI growth is driving demand for advanced optical networks, distributed computing infrastructure, and regional data center connectivity.
According to Ciena’s Wave Services Report, global revenue for high-capacity optical connectivity services is projected to grow at an average annual rate of 18%, while Southeast Asia is expected to exceed that pace with a 22% compound annual growth rate (CAGR). The growth is being driven by demand for 100G, 200G, and 400G+ connectivity services needed to support AI workloads.
Neoscalers reshape AI infrastructure landscape
A major trend discussed during the roundtable is the emergence of neoscalers—specialized providers offering GPU-as-a-service and AI infrastructure capabilities alongside traditional hyperscalers.
Unlike hyperscalers that typically operate large proprietary infrastructures, neoscalers rely on highly automated, flexible, and scalable network services to quickly deploy AI computing resources. As AI applications move from model training to real-time inference, these providers are becoming increasingly important in delivering low-latency AI services closer to users.
Global investment in AI inference infrastructure is expected to surpass AI training investments in 2026, creating demand for more geographically distributed computing facilities across Southeast Asia.
AI drives expansion of regional data center corridors
The rapid growth of AI workloads is also reshaping the region’s data center landscape. Physical limitations in established hubs, including land and power constraints in Singapore, are encouraging investments in emerging locations such as Johor, Malaysia, and Indonesia.
TM Global is expanding its hyperconnected data centers in Klang Valley and Johor to support AI training and inference workloads. The Philippines is also expected to play an important role as a subsea connectivity gateway linking international traffic with regional AI infrastructure corridors.
CSPs move toward AI-ready networks
To support AI-driven demand, communication service providers (CSPs) are shifting from traditional fiber connectivity offerings toward automation-driven, AI-ready infrastructure.
Key technologies enabling this transformation include:
- Coherent 800G optics for moving massive AI datasets with improved efficiency
- Advanced network automation for faster deployment and dynamic capacity management
- AI-powered network management for predictive maintenance and optimized performance
- Low-latency network designs optimized for GPU-based parallel processing
CSPs are also increasing physical network capacity, with TM Global noting that cable deployments are evolving from 96 fiber cores to 384 fiber cores to address growing bandwidth requirements.
“AI infrastructure players build for massive parallel processing and constant data movement. This requires a different mindset—one that is much more capacity-led, performance-led, and time-to-market driven,” said Alex Wong of Ciena.
Data sovereignty becomes a strategic priority
Beyond connectivity and computing power, data sovereignty is becoming a major factor in AI infrastructure decisions. Organizations and governments are increasingly prioritizing where data is processed, stored, and controlled.
Francisco Sant’Anna of Ciena said regional providers have an opportunity to become “Technology National Champions” by enabling sovereign AI ecosystems that keep data within local jurisdictions while allowing access to global innovation.
As Southeast Asia transitions from a connectivity market into an AI infrastructure hub, industry leaders believe the region’s future competitiveness will depend on building networks capable of supporting large-scale AI workloads, distributed computing, and secure digital ecosystems.
“Southeast Asia is no longer just a transit node—it is rapidly becoming a major global AI hub,” said Baharul Nizam Said Daliman of TM Global. “We aren’t just building networks; we are laying the digital foundation that allows AI to scale and thrive.”
