Key Takeaways

  • Qualcomm introduced a broad AI data center roadmap, targeting the rapidly growing artificial intelligence infrastructure market.
  • Goldman Sachs projects roughly $7.6 trillion in cumulative AI-related capital expenditures between 2026 and 2031 across compute, data centers, and power.
  • Hyperscalers, including Amazon, Google, Microsoft, and Meta, are expected to spend more than $350 billion on capital expenditures in 2025.

Big Tech hyperscalers and artificial intelligence leaders are driving an unprecedented capital-expenditure boom, with AI infrastructure investment now measured in the trillions. This spending surge is reshaping data center strategy, semiconductor investment, and power demand. JPMorgan Global Research's midyear outlook highlights how quickly this cycle is developing, noting that AI-driven capital spending is scaling rapidly while the underlying economics remain attractive.

At the center of the expansion is what JPMorgan calls AI upstream investment, an umbrella that includes data centers, advanced chips, and the underlying infrastructure required to support scaled AI workloads. This progression is anchored heavily in North America, as the United States currently accounts for roughly 85% of AI and machine learning venture capital. This concentration drives spillover benefits into China, South Korea, and Taiwan, particularly through their critical roles in the semiconductor supply chain.

Global research firms have documented this structural acceleration with remarkable consistency. McKinsey estimates that global data center infrastructure capital expenditures will reach almost $7 trillion by 2030, largely driven by AI workloads. Similarly, Dell'Oro Group projects that AI-related data center investments will help drive global infrastructure spending to about $1 trillion by 2028. In parallel, broader forecasts from Goldman Sachs project roughly $7.6 trillion of cumulative AI-related expenditures between 2026 and 2031 across compute, data centers, and power. Reflecting this growth, Stanford's AI Index reports $37 billion in global private investment specifically in AI infrastructure in 2024, highlighting rapid acceleration from a relatively small base.

Against that backdrop, Qualcomm recently unveiled a comprehensive AI data center strategy at its Investor Day, targeting a meaningful share of this massive wave of investment. While traditionally best known for its smartphone and mobile chip business, the company is aggressively expanding into the data center space to capture a share of the rapidly growing AI infrastructure opportunity.

The CFO and COO of Qualcomm emphasized that the company is entering a key inflection point, supported by multi-year and multi-generation customer agreements. It is common to see chip companies diversify when market cycles shift, but Qualcomm's timing puts it in direct competition with hyperscaler-aligned silicon providers and GPU specialists that dominate AI compute. Nvidia, for example, has been central to GPU-dense AI data center designs, a trend that is unlikely to subside in the near term. Qualcomm is betting that the market has room for another major competitor.

Hyperscalers continue to feature heavily in these infrastructure forecasts. The four largest hyperscalers, Amazon, Google, Microsoft, and Meta, are expected to spend more than $350 billion on capital expenditures in 2025. Broader technology players are pushing total industry capital expenditures toward $500 billion, illustrating how rapidly balance sheets are evolving to accommodate new compute demands.

Even with rising demand for AI services, companies are evaluating whether enterprise monetization will scale at the same pace as capital expenditure. Revenue from AI-powered products is rising, but long-term enterprise adoption must accelerate to justify current spending trajectories. To manage the complexities of this deployment phase, organizations are increasingly looking to established frameworks, such as IEEE standards for data center power and networking, alongside ISO/IEC AI governance and risk-management standards.

The AI infrastructure cycle remains in full acceleration. Qualcomm's new data center strategy adds a formidable competitor to the rapidly expanding field. Hyperscaler spending continues to rise sharply, and industry analysts from McKinsey to Dell'Oro Group are tracking capital flows that will dictate where compute and networking capacity will concentrate for years to come. The technology sector remains committed to building for a future of scaled AI workloads at unprecedented levels.