Southeast Asia’s digital infrastructure is undergoing a fundamental shift. Enterprises are facing a combination of pressures that can no longer be addressed through incremental upgrades. This includes security systems that take years to change, AI workloads that push networks beyond their original design, data sovereignty rules and energy limits that restrict how much infrastructure can grow.
As a result, network architecture decisions are no longer routine operational choices, but are long-term commitments that affect security, performance, compliance, and cost across the region.
The pressure is most immediate in security. While quantum computing is often discussed as a future concern, the systems it threatens are deeply embedded and slow to change.
Realistically, enterprises across Southeast Asia should begin planning and small-scale testing of PQC (post-quantum cryptography) within the next 12 to 18 months, ahead of early production adoption between 2028 and 2029, Colt Technology Services’ APAC president, Yasutaka Mizutani told iTnews Asia.
Security and sovereignty are driving architectural change
Migrating cryptographic systems across sectors such as banking, aviation and manufacturing can take years, and by the time fault-tolerant quantum computers become commercially viable in the early 2030s, organisations that have not begun the transition may find themselves exposed, said Mizutani.
He added that this timeline is reinforced by developments in Singapore, where the National Quantum Office and US-based firm Quantinuum will set up the Helios quantum computer for commercial use in 2026.
The project aims to help local companies explore practical quantum applications in areas of drug discovery, financial modelling and logistics optimisation, signalling that quantum computing is moving from theory to commercial reality faster than many anticipated.
This underlines the urgency for enterprises to prepare their security systems for a quantum-ready world, said Mizutani.
Security considerations do not stop at encryption.
Across Southeast Asia, enterprises operate under diverse regulatory regimes that define where data can reside and how it can move.
Data sovereignty remains a key consideration across multiple jurisdictions in Southeast Asia, said Mizutani.
“Each country has its own regulatory framework, and our approach is to align infrastructure deployment with local requirements while maintaining seamless regional connectivity.
“By keeping data within the boundaries where it originates, we respect sovereignty laws while still enabling cross-border collaboration,” he said.
For latency-sensitive industries, including financial services, aviation, and smart manufacturing, performance requirements now extend into microseconds. The ability to process and transmit data at ultra-low latency is critical to maintaining competitiveness.
To achieve this, the focus should be on building a distributed network architecture that places edge nodes strategically close to demand centres, while optimising routes to cloud-based infrastructure, said Mizutani.
This ensures that data is acted upon in real time, reducing latency to the absolute minimum, he added.
Furthermore, AI workloads are placing pressure on existing data centre and network models.
AI is forcing a rethink of network topology and operations
Training and inference generate sustained, high-volume data flows that legacy, centralised architectures were not designed to handle efficiently. As data-centre capacity across Southeast Asia continues to expand rapidly, it has become clear that scale alone is not the answer.
According to Mizutani, the need for decentralisation has become evident.
Enterprises should adopt a more distributed model where data is processed closer to its source, reducing backhaul traffic and latency. This does not mean replacing core data centres, but complementing them with smaller, edge-based facilities that handle inference and local processing.
- Yasutaka Mizutani, President, APAC, Colt Technology Services
The backbone network connecting these nodes should use optical technologies and software-defined orchestration to dynamically route traffic where capacity is needed most, he added.
This hybrid architecture improves resilience, reduces operational costs, and supports real-time AI applications including autonomous systems and predictive analytics across industries.
“Operationally, this shift is being reinforced by AI-powered orchestration and predictive analytics.
“By continuously monitoring telemetry across optical and IP layers, networks can detect early signs of congestion or failure and respond before performance degrades,” said Mizutani.
He added that integrations such as those between Colt’s IQ Network and Oracle Cloud Infrastructure illustrate how on-demand, software-defined connectivity can be enhanced through automation.
Energy constraints and collaboration will define what scales
Data growth across Southeast Asia is outpacing available energy capacity, turning efficiency into a hard limit on expansion.
To address this, operators are increasingly turning to next-generation optical transport and photonic technologies that deliver higher bandwidth at lower power per bit, said Mizutani.
He added that optical switching, coherent transmission and photonic-integrated circuits are improving efficiency while enabling massive capacity expansion.
Equally, data centre design is evolving to reduce overall power usage effectiveness through liquid cooling, AI-assisted load management, and the integration of renewable energy.
Mizutani said locating facilities in climates that allow free cooling and integrating edge processing to reduce redundant data transport also play a major role.
“The industry is moving toward a model where every transmitted bit has a measurable energy footprint, making energy-per-bit a key metric in network and data-centre planning,” he said.
Sustainability is becoming synonymous with performance, he added.
However, no single enterprise or service provider can address these challenges alone.
According to Mizutani, building a regional digital backbone that supports AI, security and sustainability will require close cooperation between governments, carriers and large enterprises.
“Governments must lead by establishing interoperable standards for encryption, latency and energy performance, while service providers and enterprises jointly invest in shared fibre routes, regional data-exchange points and quantum-safe pilot networks,” he said.
