Osipov Vladislav

Vladislav Osipov

Nvidia will be one of the main beneficiaries of hyperscaler cost growth totaling $660 billion in 2026 / Photo: Mijansk786/ Shutterstock.com

Nvidia will be one of the main beneficiaries of hyperscaler cost growth totaling $660 billion in 2026 / Photo: Mijansk786/ Shutterstock.com

Quotes for chipmaker Nvidia jumped nearly 8% on Friday, February 6. That's the strongest increase in nine months, according to Dow Jones Market Data cited by MarketWatch. Investors have started actively buying up securities of companies that could benefit from the colossal spending that tech giants have promised to devote this year to infrastructure for artificial intelligence. Friday's rally ended Nvidia's longest losing streak since Sept. 3, when the stock fell in price for five consecutive trading sessions due to a general market sell-off.

Details

Nvidia shares rose 7.8% in trading Feb. 6, the most since April 9, 2025, MarketWatch calculated. The company's market value increased by about $325 billion, the fourth-largest one-day capitalization gain in history for an individual stock, Bloomberg reports.

The recovery was spurred by Amazon saying during a conference call after the release of its quarterly report that it plans to increase capital spending this year and invest $200 billion in data centers, chips and other equipment. That's significantly more than Wall Street had envisioned and than competitors have claimed. Alphabet will also increase investments, they will make from $175 billion to $185 billion. Meta's expenses may reach from $115 billion to $135 billion. Microsoft did not report on plans for capex, but in the previous quarter they increased in annualized terms by 66%.

Combined, these hyperscalers could funnel as much as $660 billion in investments this year, with much of the money going to buy Nvidia chips, CNBC writes.

The tech industry's rapid growth in capital spending on AI infrastructure is justified, reasonable and sustainable Nvidia CEO Jensen Huang told CNBC on Friday. "The reason is that all of these companies will start to have growing cash flow," he explained. Huang called what's happening "the largest infrastructure build in human history," which is being driven by "incredible" demand for computing power. According to the Nvidia chief, AI companies and cloud platforms can use this power to increase their profits.

Who else got the impetus

After three days of decline, the technology sector showed growth in trading on Friday. The securities of companies related to AI equipment were especially active. In addition to Nvidia, shares of chipmakers Broadcom and Advanced Micro Devices also jumped by 8%.

Quotes of Super Micro Computer, which assembles servers based on Nvidia chips, soared by almost 12. The capitalization of the manufacturer of components for data centers Astera Labs increased by 18%. The securities of cloud services provider CoreWeave rose about 20%. Data center power equipment companies including Amphenol, GE Vernova and Vertiv Holdings also gained in value.

What the analysts are saying

"I think we're seeing investors locking in profits on Nvidia, which they've probably already made a good profit on," Melissa Otto, head of Visible Alpha research at S&P Global, told MarketWatch. Investors may switch to stocks of memory makers that are also benefiting from the investment boom but are also showing explosive growth in earnings forecasts - similar to what happened with Nvidia a few years ago, she said.

Digital data storage companies Sandisk, Micron Technology and Western Digital are also falling into this trend, MarketWatch points out. Their stocks also rose sharply on Friday and remain among the top gainers in the S&P 500 this year.

According to Jordan Klein, an analyst at Mizuho's trading division, the market is realizing that demand for AI computing and data center investments will intensify not only in 2026, but also in 2027. The race to increase capex among tech giants will only accelerate, he believes, as AI is deployed more widely - in applications and products that will require even more resources for inferencing (the process of executing AI models). Even though supply of some key components for data centers is growing, Klein believes there won't be enough to meet demand, especially in light of large orders beyond 2026.

This article was AI-translated and verified by a human editor

Share