Artificial intelligence is entering a more infrastructure-driven phase, as companies at GITEX Asia 2026 highlighted growing constraints around compute, energy and hardware supply during the event's opening in Singapore.
The event, held at Marina Bay Sands from April 9–10, has drawn participants from more than 110 countries, including over 550 enterprises and startups and 250 global investors managing about US$350 billion in assets, underscoring its scale as one of the region's largest AI gatherings.
From models to deployment
Across the exhibition floor and conference agenda, companies are shifting their focus from model development to how AI can be deployed at scale, according to industry participants.
The shift comes as constraints around compute resources, energy consumption and hardware availability increasingly shape how AI systems are commercialized.
Companies are placing greater emphasis on efficiency and deployment models that can operate across both edge and data center environments, particularly in industrial and enterprise applications.
Infrastructure emerges as bottleneck
Executives said infrastructure is emerging as a key constraint, as demand for AI computing drives rapid expansion in data center capacity across Singapore and neighboring markets.
"The challenge now is less about building models and more about whether infrastructure can support deployment at scale," said Chris McCuin, managing director for Southeast Asia at Kuan International.
He added that limitations in power, cooling and hardware availability are increasingly shaping how and where AI systems can be deployed.
The buildout is enabling more complex architectures, including hybrid AI models that balance edge inference with centralized processing, though many deployments remain in validation or early-stage rollout, industry participants said.
Companies such as Blaize, working with Nokia, are showcasing integrated solutions developed in Singapore aimed at improving latency and energy efficiency in real-world environments.
Energy management and cloud capacity are becoming critical factors as enterprises seek to scale AI while managing cost and sustainability constraints.
Southeast Asia gains strategic importance
Singapore is positioning itself as a regional hub linking Asia's technology ecosystem with global capital and markets, as companies expand across Southeast Asia and seek cross-border partnerships.
The city ranked fourth globally in the 2025 Global Startup Ecosystem Index, reflecting its role in enterprise and deep-tech innovation.
Participants said Southeast Asia is gaining traction as companies localize infrastructure and adjust to evolving supply chain conditions, particularly amid continued constraints in access to advanced hardware.
The presence of both Chinese and international firms at the event underscores a broader shift in the AI industry, where efficiency, deployment and infrastructure are becoming more critical than model scale alone.
Rather than focusing solely on training larger models, companies are prioritizing optimization, energy efficiency and faster commercialization cycles.
The trend points to a new phase in AI development, where access to compute, power and deployment capability is shaping competitive advantage.
Article edited by Jack Wu
