Analyst Spotlight
Matt Eastwood
SVP, WW Research
IDC
AI Infrastructure: The Foundation of the Agentic Era
The world is entering a defining moment for digital infrastructure. Artificial intelligence has moved from experimentation to ubiquity, and with it, a new operational paradigm is taking shape — one where agents rather than applications become the primary engines of digital value creation. This is the dawn of the agentic AI era, and its success depends on one thing above all else: robust, intelligent, and scalable infrastructure.
From Automation to Autonomy
For decades, infrastructure strategy has focused on efficiency – making IT faster, cheaper, and more reliable. But AI is forcing a step change. IDC’s Worldwide IT Industry 2026 FutureScape predicts that by 2028, nearly half of all IT product and service interactions will be mediated by AI agents. These systems are not just automating tasks; they are reasoning, collaborating, and acting in context – continuously learning from data to improve business outcomes.
Supporting this shift requires infrastructure that can think for itself. IDC’s Future of Digital Infrastructure research shows that by 2029, 70% of new operating systems will ship with built-in infrastructure operations agents and model context servers to drive efficiency, security, and sustainability. In short, we are moving from systems that are operated to systems that operate themselves.
AI Factories and the Rise of Private Intelligence
The massive growth of generative and agentic AI has triggered a global infrastructure renaissance. Enterprises and hyperscalers alike are building “AI factories”. These are the next-generation data centers purpose-built for high-density and GPU-driven workloads. AI-ready data center spending in the U.S. has tripled in three years and forecast anticipate that demand for AI-ready capacity will grow 33% annually through 2030.
IDC’s recent Private AI Infrastructure Systems MarketScape underscores why this matters: as AI workloads scale, organizations need hybrid models that balance performance, cost, and control. Leaders like Dell Technologies, HPE, and Cisco are responding with turnkey private AI systems that integrate compute, storage, networking, and model management software into secure, cloud-consistent platforms. These systems form the backbone of enterprise AI, where data sovereignty, security, and latency matter most.
The Power, Cooling, and Connectivity Challenge
The scale of AI infrastructure buildout is also testing physical limits. High-density GPU clusters can draw tens of kilowatts per rack, driving record levels of power demand and forcing innovation in liquid cooling and grid optimization. IDC predicts that by 2030, 70% of new liquid-cooled deployments will adhere to open standards, improving compatibility and reducing deployment costs by one-third. The infrastructure bottleneck is shifting from compute to power and cooling, making sustainability not just an ESG issue but an operational imperative.
Toward the Autonomous Enterprise
Agentic AI doesn’t live in isolation – it depends on a digital fabric that spans datacenters, clouds, and edge environments. By 2027, IDC expects 80% of enterprises to deploy distributed edge infrastructure to support low-latency AI inferencing, and 75% will use interconnection-oriented networks to secure and orchestrate AI workloads. This fusion of automation, intelligence, and interconnection is paving the way toward autonomous IT operations, where humans remain in the loop but not in the way.
Why It Matters Now
CIOs in ASEAN and beyond are standing at the intersection of two transformations: the modernization of infrastructure and the emergence of the agentic enterprise. The winners will be those who view AI infrastructure not as a cost center but as a catalyst – the intelligent backbone that allows agents, data, and humans to collaborate seamlessly.
At the IDC ASEAN AI & Data Summit, we’ll explore how forward-thinking leaders are reimagining infrastructure for this new era by building the secure, sustainable, and scalable foundations of an intelligent enterprise. Because in the age of agentic AI, infrastructure isn’t just the platform for innovation. It is the innovation.
