GITEX Asia 2026 wrapped up in Singapore with a clear shift in industry priorities, as artificial intelligence (AI) spending moves from infrastructure buildout toward monetization, with inference and edge deployment emerging as the next battleground.
The two-day event, held April 9-10 at Marina Bay Sands, drew participants from more than 110 countries, including over 550 enterprises and startups, alongside investors managing about US$350 billion in assets, underscoring Southeast Asia's growing role in the global AI ecosystem.
From buildout to deployment
Discussions across the exhibition floor moved beyond model development toward the operational realities of scaling AI in production environments.
Constraints in compute availability, energy consumption, and hardware supply are increasingly shaping how companies commercialize AI, pushing the industry to focus on efficiency and deployability rather than model scale alone.
Companies are prioritizing architectures that can operate across both edge and data center environments, particularly in enterprise and industrial use cases where performance and cost efficiency are critical.
Infrastructure meets physical limits
The push toward large-scale deployment is driving rapid expansion of data center capacity across Singapore and neighboring markets, but executives said infrastructure constraints are becoming more pronounced.
Power availability, cooling requirements, and access to advanced hardware are emerging as key bottlenecks that could limit growth.
These constraints are also reshaping system design. Companies are adopting hybrid architectures that distribute workloads between centralized data centers and edge environments to reduce latency and optimize resource use.
Nokia and Blaize demonstrated joint solutions developed in Singapore, integrating networking infrastructure with AI inference platforms for real-world deployment.
Monetization shifts to inference
As infrastructure investment scales, attention is shifting toward how AI generates returns.
Executives said the next phase of growth will depend less on building larger models and more on deploying AI in practical applications that deliver measurable business outcomes.
"What we're seeing now is that investment is shifting toward inference and the edge, where the next wave of monetization is expected," said Stephen Patak.
He added that while training remains essential, revenue generation is expected to come from deployment rather than model development.
"The real revenue is not coming from training, but from real-world use cases at the edge," Patak said.
Industry participants said inference remains in an early stage but is expected to expand as improvements in power efficiency, cost, and system performance enable broader adoption across sectors.
Southeast Asia gains strategic weight
The event highlighted Southeast Asia's growing importance as companies localize infrastructure and adapt to shifting supply chain dynamics.
Singapore is positioning itself as a regional hub linking global technology providers with capital and enterprise demand, reinforcing its role in the AI value chain.
The presence of both Chinese and international firms reflects a broader shift in the industry, where deployment capability, efficiency, and infrastructure resilience are becoming more important than raw model performance.
Rather than focusing solely on training larger models, companies are prioritizing optimization, cost control, and scalable deployment using available hardware.
As GITEX Asia 2026 concluded, the direction of the industry became more defined. The AI race is no longer driven solely by model scale or infrastructure expansion, but by how effectively systems can be deployed, monetized, and translated into real-world applications.
Article edited by Jack Wu
