NextFin News - In a high-stakes defense of the current technological trajectory, Nvidia Corp. Chief Executive Officer Jensen Huang declared on Friday, February 6, 2026, that the massive capital expenditures currently flowing into artificial intelligence (AI) infrastructure are not only appropriate but sustainable for the long term. Speaking in an interview with CNBC, Huang addressed growing investor anxiety regarding the spending levels of "hyperscalers" like Microsoft, Alphabet, and Amazon, asserting that the build-out of AI data centers will continue for at least another seven to eight years. Huang emphasized that global demand for AI remains "incredibly high," dismissing fears of a speculative bubble by pointing to the fundamental replacement of the world’s existing $1 trillion data center install base with accelerated computing architectures.
The timing of these remarks is critical, as the market has shown signs of fragility. Just 24 hours prior, Amazon reported that its capital expenditures could reach $200 billion in 2026—a 50% increase over the previous year—which initially triggered a 10% slide in its stock price. However, the market found a floor following Huang’s comments and a reaffirmation from Goldman Sachs, which maintained a "Buy" rating on Nvidia with a $250 price target. According to analyst James Schneider at Goldman Sachs, the bank expects favorable supply-and-demand trends to persist into 2027, bolstered by the successful rollout of Nvidia’s Blackwell and upcoming Rubin architectures. This sentiment was echoed by Simon Lin, chairman of Nvidia supplier Wistron, who predicted that AI-related orders in 2026 would exceed the growth seen in 2025.
From a senior financial perspective, Huang’s assertion rests on the concept of the "AI Capital Expenditure Supercycle." We are witnessing a structural shift where computational power is being treated as a sovereign resource, akin to oil or gold. Under the current U.S. President Trump administration, the focus on domestic industrial strength has aligned with this "physicality" shift. Analysts at Barclays estimate that this transition from software-based services to heavy hardware infrastructure is now contributing approximately 1% to total U.S. GDP growth. The $4 trillion figure, once a bold projection by Huang in late 2024, has now become the consensus target for global data center spending by 2030, as validated by recent reports from McKinsey & Company.
However, the sustainability of this spending faces a looming challenge known as the "Power Wall." As the industry races to build "AI factories," the primary bottleneck has shifted from chip throughput to the availability of electricity and advanced memory. Data from the PJM Interconnection, the largest grid operator in the U.S., shows that capacity auction prices surged by over 800% in late 2025, with a significant portion of that increase attributed to data center demand. This has created a new class of market winners: utility companies like Constellation Energy and Vistra Corp., which are signing "behind-the-meter" deals to provide dedicated nuclear power to AI campuses. Huang’s confidence suggests that the return on investment (ROI) for AI will be found in these efficiency gains and the creation of new industrial capabilities that justify the high energy costs.
Looking forward, the trajectory of AI spending will likely be defined by "energy autonomy." To maintain the sustainability Huang describes, tech giants are expected to pivot toward becoming power producers themselves, investing heavily in Small Modular Reactors (SMRs) and massive battery storage. The next two years will be a period of "forced growth," where the limits are physical—land, power, and specialized packaging like TSMC’s CoWoS. While the risk of a sentiment shift remains if quarterly earnings from hyperscalers fail to show direct AI monetization, the underlying replacement of global computing infrastructure provides a multi-year floor for demand. As Huang noted, the world is not just building servers; it is building a new class of industrial infrastructure that will underpin the global economy for the next decade.
Explore more exclusive insights at nextfin.ai.