New York, Feb 9, 2026, 10:36 (EST) — Regular session.
- Nvidia picked up 3.5%. AMD and Broadcom pushed more than 2% higher during the morning session.
- Micron slipped 1.3%, out of step with the broader chip rally, as traders digested changing demand trends for next-gen AI memory.
- Traders are eyeing Big Tech’s AI investments, looking for signs of payoff, with Nvidia’s remarks at GTC in March drawing particular attention.
Nvidia jumped 3.5% on Monday, pacing a rally in AI chip names. Micron lost 1.3%, diverging from a lift in broader tech and semiconductor ETFs. AMD put up a 2.5% gain. Broadcom was up 2.4%. Microsoft edged higher by 2.3%, with Palantir out front, up 4.5%.
Chipmakers caught another wave of buying after a choppy period for AI-related stocks, with investors still debating the timeline for these massive investments to start boosting profits. Reuters flagged that the top four U.S. tech giants are expected to spend roughly $650 billion on capital projects this year. Next up, traders are zeroing in on upcoming U.S. jobs, inflation, and consumer spending figures to gauge the outlook for rates and risk.
AI stocks aren’t moving in lockstep anymore. Reuters data points to a split: chip and data-center names are increasingly seen as the winners—investors are lumping them together as the classic “picks ‘n shovels.” Software and data stocks, on the other hand, are getting hit, with the market betting they’re at risk of being upended by AI. “This divergence is … a signal that investors are differentiating between who enables AI and who may be disrupted by it,” Saxo’s Charu Chanana wrote. Mark Hawtin at Liontrust put it bluntly: the market isn’t interested in companies that spend just for the sake of it. The Journal Record
Monday saw chip-related stocks catch a fresh wave of attention on the back of high-bandwidth memory, or HBM—those vertically stacked chips powering AI servers. According to TrendForce, Samsung is gearing up to start mass production and shipping of next-gen HBM4 right after the Lunar New Year. Citing Yonhap and Hankyung, the report says shipments could begin as soon as the third week of February, pending Nvidia quality sign-off. Still, TrendForce pointed out that Samsung’s decision to use more advanced manufacturing processes comes with higher production costs and some yield uncertainty.
Memory isn’t just background noise for Nvidia. HBM now acts as both a bottleneck and a key expense as chipmakers race to deliver speedier accelerators for data centers. Even minor changes in the supplier landscape can swing sentiment fast.
Micron faces a separate challenge as investors weigh its position and timing in HBM4, a segment where Korean competitors are pushing hard for first-mover allocations. If Micron appears behind in capturing a platform cycle, expectations can slip, regardless of how the sector is trading overall.
But the downside risk isn’t off the table. Should Big Tech pull back on spending, or if higher rates keep squeezing long-duration growth stocks, chip stocks could quickly lose ground. HBM4 rollouts running behind schedule or costing more than anticipated could also put margin pressure on the so-called “AI hardware” winners.
Nvidia’s GTC conference, set for March 16–19 in San Jose, is the next key date circled by the group. Investors are watching closely for updates on the product pipeline and how fast the AI data-center buildout is moving.
