NEW YORK, February 9, 2026, 08:01 EST — Premarket
- Nvidia shares dip in premarket trading after Friday’s sharp rebound
- Traders are weighing Big Tech AI spending plans ahead of Nvidia’s late-February results
- U.S. jobs and inflation reports later this week could swing rate bets and chip stocks
Nvidia shares were down 0.9% at $183.70 in premarket trade as of 7:37 a.m. ET, easing after a 7.9% jump on Friday that ended at $185.41. The stock traded between $174.60 and $187.00 in the prior session, with about 231 million shares changing hands.
The pullback came as investors leaned back from the AI trade again, after a run of earnings updates from Big Tech revived worries over capital expenditure — cash spent on items like data centers, servers and chips. Amazon, Google, Meta and Microsoft are expected to spend about $650 billion as they push deeper into AI, Reuters reported.
“After a hectic week last week,” the size of Friday’s rebound “didn’t feel like the beginning of a sustainable reversal,” said Ipek Ozkardeskaya, senior market analyst at Swissquote Bank. Futures were lower and other chip names also slipped before the open, while traders waited for economic data that has been pushed back by a partial U.S. government shutdown. Kitco
Friday’s rally in Nvidia was helped by a fresh round of AI hardware spending talk from its biggest customers, and by comments from CEO Jensen Huang. Huang told CNBC that demand has been “going through the roof,” and said AI could be reaching an “inflection point,” according to Investopedia. Investopedia
Separate reports out of South Korea added another supply-chain angle for the sector. Korea JoongAng Daily, citing Yonhap and industry sources, said Samsung Electronics plans to begin mass production of HBM4 — high-bandwidth memory used alongside GPUs — later this month and ship as early as next week for use in Nvidia GPUs. The report said Samsung has passed Nvidia’s quality certification and secured purchase orders, with production timing aligned to Nvidia’s next-generation “Vera Rubin” AI accelerator plans. Korea Joongang Daily
High-bandwidth memory is a stacked, high-speed component that sits close to the processor and can be a bottleneck for supply and margins in advanced AI systems. Any shift in who can qualify and ship at scale can matter, especially as customers push for more supply options.
Investors’ next hard catalyst on Nvidia itself is earnings. The company is scheduled to report fourth-quarter fiscal 2026 results on February 25 at 2:00 p.m. PT, according to its investor events calendar.
Beyond the headline numbers, traders are likely to focus on the tone of demand commentary from cloud and data-center customers, and on whether Nvidia flags any new supply constraints or pricing pressure as competition in AI hardware heats up.
But the week also carries macro risk. If the delayed jobs report or CPI prints hotter than expected, it could push yields higher and weigh on high-growth stocks, including semiconductors, even if the AI demand story stays intact.
For now, Nvidia is trading like a macro-sensitive bellwether again: one part earnings story, one part rate story. Markets will watch Fed speakers on Monday, then the delayed January nonfarm payrolls report on Wednesday and January CPI on Friday — with Nvidia’s February 25 results looming as the next company-specific test.
