Nvidia shares rose 7% during Friday trading after CEO Jensen Huang told CNBC’s “Halftime Report” that the tech industry’s massive $660 billion capital expenditure surge for AI infrastructure is not only justified but sustainable. Huang emphasized that the current wave of spending is driven by strong demand for computing power and the clear path to future profitability. His remarks followed recent earnings reports from major Nvidia clients—Meta, Amazon, Google, and Microsoft—all of which announced plans to significantly increase investment in AI infrastructure. Collectively, these hyperscalers could spend up to $660 billion this year, with a substantial portion allocated to purchasing Nvidia’s GPUs. While Wall Street reacted mixedly—with Meta and Alphabet seeing gains, and Amazon and Microsoft’s stocks declining—Huang remained confident in the long-term economic logic behind the spending. He described the current buildout as “the largest infrastructure expansion in human history,” fueled by “sky-high” demand for AI-driven computing. Huang highlighted concrete examples of how customers are deploying AI. Meta, he said, is transitioning from traditional CPU-based recommendation systems to advanced generative AI and autonomous agents. Amazon Web Services is leveraging Nvidia chips to enhance product recommendations and cloud-based AI services. Microsoft, he noted, is using Nvidia-powered AI to boost productivity and intelligence in its enterprise software offerings. He also praised leading AI labs OpenAI and Anthropic, both of which rely on Nvidia’s hardware through cloud providers. Nvidia invested $10 billion in Anthropic last year, and Huang confirmed the company plans to make a major investment in OpenAI’s upcoming fundraising round. “Anthropic is making great money. OpenAI is making great money,” Huang said. “If they could have twice as much compute, their revenues would go up four times as much.” He pointed to the continued utilization of older Nvidia chips—such as the A100, released six years ago—as evidence of sustained demand. Even these legacy chips are now being rented out at scale, indicating a robust and ongoing need for AI compute. “Once people start paying for AI services and those companies generate real profits, the cycle begins to accelerate,” Huang said. “They’re going to keep doubling, doubling, doubling, doubling.”