NextFin News - Nvidia Chief Financial Officer Colette Kress has projected that annual global spending on artificial intelligence infrastructure could reach between $3 trillion and $4 trillion by the end of the decade. Speaking during the company’s May 20 earnings call, Kress characterized AI as a "necessity" rather than a luxury, signaling a fundamental shift in how corporations and sovereign nations view high-performance computing. The forecast comes as Nvidia’s market valuation surpassed the $5 trillion mark in early 2026, fueled by what CEO Jensen Huang described as "parabolic" demand for the company’s hardware and software ecosystem.
Kress, who has served as Nvidia’s CFO since 2013 and is currently ranked among the most powerful women in business by Fortune, has long maintained a bullish stance on the structural transition toward accelerated computing. Her tenure has seen Nvidia evolve from a niche graphics chipmaker into the primary architect of the AI era. While her projections are often viewed as a bellwether for the broader technology sector, they represent the outlook of a company that currently controls over 80% of the AI chip market. Consequently, these figures reflect Nvidia’s internal modeling of total addressable market expansion rather than a consensus figure verified by independent economic agencies.
The $4 trillion figure represents a massive escalation from current global IT spending levels. According to Kress, this capital expenditure will be driven by a diverse group of buyers, including "hyperscalers" like Microsoft and Amazon, sovereign AI initiatives where nations build their own domestic computing capacity, and the rise of "agentic AI." This newer form of artificial intelligence, which can perform complex, multi-step tasks autonomously, is cited by Huang as the primary catalyst for the current "parabolic" demand curve. Huang noted that in the current economic landscape, "compute capacity is revenue and profits," suggesting that the bottleneck for global economic growth has shifted from labor or capital to the availability of silicon.
However, the scale of this projected spending has met with skepticism from some corners of the financial community. Analysts at several major investment banks have raised concerns about the "return on investment" (ROI) gap, questioning whether the software revenue generated by AI applications can eventually justify the trillions of dollars being poured into hardware. While Nvidia’s revenue has soared, the broader enterprise market is still in the early stages of proving that AI can deliver the productivity gains necessary to sustain such high levels of capital investment. If these productivity gains fail to materialize at scale, the $4 trillion annual spending target could prove to be an optimistic scenario rather than a market certainty.
The geopolitical dimension of this spending also introduces significant variables. Sovereign AI initiatives, which Kress highlighted as a key growth driver, are heavily influenced by national security policies and trade restrictions. As U.S. President Trump’s administration continues to navigate complex trade relations, particularly regarding high-end semiconductor exports, the ability of Nvidia to access a truly global $4 trillion market remains subject to regulatory shifts. The concentration of AI infrastructure spending in a few major geographic hubs and among a handful of corporate giants suggests that the path to Kress’s projected total will be as much about political stability as it is about technological innovation.
Explore more exclusive insights at nextfin.ai.
