AI Boom Propels Global Data Center Capital Expenditure Toward $1.7 Trillion by 2030

The global infrastructure landscape is undergoing a seismic shift as the artificial intelligence revolution forces a complete recalibration of capital allocation. According to a new forecast from the Dell'Oro Group released on February 11, 2026, the unrelenting demand for AI capabilities is projected to drive worldwide data center capital expenditure (capex) to a staggering $1.7 trillion by 2030.

This forecast marks a significant upward revision in industry expectations, signaling that the initial hype surrounding generative AI has transitioned into a sustained, multi-year infrastructure expansion cycle. For industry observers and stakeholders, the report underscores a critical reality: the physical backbone of the internet is being rebuilt from the ground up to accommodate the unique power, cooling, and compute requirements of next-generation AI models.

The Trillion-Dollar Milestone Arrives Early

One of the most striking findings in the Dell'Oro report is the acceleration of near-term spending. Global data center capex is now expected to approach the $1 trillion milestone in 2026, a threshold reached significantly sooner than previous models anticipated. This acceleration is not merely a result of incremental growth but a fundamental step-change in how digital infrastructure is deployed.

Baron Fung, Senior Research Director at Dell'Oro Group, noted that the industry is entering a new phase of expansion. "Hyperscale and neo cloud service providers, along with sovereign AI initiatives, are entering a new phase of infrastructure expansion," Fung stated. This phase is characterized by a departure from general-purpose computing toward highly specialized, capital-intensive AI clusters.

Hyperscalers Lead the Charge with $600 Billion War Chest

The driving force behind this unprecedented spending spree remains the "Big Four" US hyperscalers: Amazon, Google, Meta, and Microsoft. These tech giants have collectively entered 2026 with aggressive momentum, raising their combined data center capital expenditures to nearly $600 billion.

Despite growing market scrutiny regarding the return on investment (ROI) for AI infrastructure, these companies are doubling down. Their strategy appears to be defensive as much as it is offensive; possessing the most robust infrastructure is seen as a prerequisite for retaining market share in an AI-defined future.

"Despite increased scrutiny around AI infrastructure returns, hyperscalers continue to invest aggressively, supported by large cash reserves and a long-term focus on market share," Fung explained.

The investment is not just about buying more servers; it is about deploying "larger and more complex AI clusters." These clusters demand a holistic upgrade of the data center environment, driving demand for:

• High-performance networking: To minimize latency between thousands of GPUs.
• Advanced storage: To handle the massive datasets required for training.
• Inference capacity: To serve models to millions of end-users.
• Advanced power and cooling: Including liquid cooling solutions necessary for high-density racks.

The Rise of Accelerated Computing

The composition of data center spending is changing as dramatically as the total volume. The Dell'Oro forecast highlights a decisive pivot toward accelerated computing. By 2030, accelerated servers—those equipped with GPUs or custom AI accelerators for training and domain-specific workloads—are expected to account for approximately two-thirds of total data center infrastructure spending.

This statistic reveals a bifurcation in the hardware market. Traditional general-purpose servers (CPUs), while still necessary, are ceding their dominance to accelerated systems designed specifically for matrix math and parallel processing. For hardware vendors, this shift dictates where R&D budgets must be focused.

Neo Clouds and Sovereign AI: The New Contenders

While the top US hyperscalers are projected to control about half of the global data center capex by 2030, the remaining market share is being contested by a new breed of players.

1. Neo Cloud Providers: specialized cloud firms focusing exclusively on AI workloads, offering bare-metal GPU access to developers.
2. Sovereign AI Initiatives: Nations investing billions to build domestic AI infrastructure to ensure data sovereignty and economic competitiveness.
3. AI Model Builders: Companies building massive proprietary clusters to train frontier models.

These groups are accelerating their deployments, contributing to the rapid approach of the $1 trillion annual spending mark.

Market Bifurcation: Enterprise Constraints

In stark contrast to the hyperscale boom, the traditional enterprise segment faces significant headwinds. The report indicates that outside of the hyperscale and neo-cloud elite, enterprise data center investment remains constrained. Factors such as tariffs, tightening monetary policy, and lingering uncertainty regarding the tangible business returns of AI adoption are causing CIOs to tread carefully. This creates a "haves and have-nots" dynamic in the industry, where cutting-edge infrastructure is increasingly concentrated in the hands of a few dominant platforms and specialized providers.

Key Forecast Data at a Glance

The following table summarizes the critical projections from the Dell'Oro Group's January 2026 Forecast Report.

Conclusion

The Dell'Oro Group's forecast serves as a barometer for the health and direction of the AI industry. The projection of $1.7 trillion in annual capex by 2030 suggests that the "AI Boom" is not a temporary bubble but a structural reorganization of the global technology stack. As hyperscalers and new entrants alike race to secure the physical assets required for intelligence, the data center is cementing its status as the factory of the 21st century. However, the widening gap between hyperscale capabilities and enterprise constraints remains a critical narrative to watch as the decade progresses.