NEW YORK, June 9, 2026, 05:07 (EDT)
Marvell Technology traded higher before the U.S. open on Tuesday, building on gains after S&P Dow Jones Indices said it would add the AI infrastructure chipmaker to the S&P 500.
The stock traded at $288.85, up around 9.6% from the last close. That puts the company’s market cap near $258 billion. Nvidia and Broadcom, both seen as AI-linked plays, also opened up in early trading.
Index changes force passive funds to buy stocks if those stocks join the benchmark. Marvell is set to move into the S&P 500 before the open on June 22, taking the spot held by Pool Corp. Flex will enter too, replacing The Campbell’s Company.
Marvell is getting another bump, this time on a mechanical move. The chipmaker had already picked up steam as an AI winner, with shares up roughly 59% since May 27. That was even after dropping 16.7% in Friday’s chip rout, according to Reuters.
Investors are betting on chips and networking equipment for AI data centers. Custom chips—built for a client’s own workloads—are now a main focus as cloud firms try to shift away from Nvidia’s pricey and often hard-to-get processors.
Marvell and Broadcom are both chip designers making custom products for cloud data centers. Marvell finds itself in a market dominated by Nvidia, but big tech buyers want more options.
Marvell reported first-quarter fiscal 2027 revenue of $2.418 billion on May 27, a company record. It sees second-quarter revenue coming in at $2.7 billion. CEO Matt Murphy pointed to “exceptional AI-related bookings” and said Marvell expects growth to keep accelerating through fiscal 2027. Net income on a GAAP basis came in at $34.5 million for the quarter. Marvell Technology, Inc.
Nvidia got in on the action, too. Jensen Huang, the chipmaker’s founder and CEO, called Marvell the next “trillion-dollar company” last week at Computex in Taipei, according to Reuters. The comment sent Marvell shares to a new high. Reuters
Nvidia and Marvell’s partnership goes deeper than optics. In March, the firms said Nvidia put $2 billion into Marvell, and the two plan to collaborate on Nvidia’s AI infrastructure setup. That includes custom XPUs and networking. XPU covers a group of specialized chips used with or in place of regular CPUs and GPUs.
Marvell is pushing to show it’s more than a custom-compute name. The company rolled out its Teralynx T100 switch silicon on June 1, claiming its 102.4-terabit-per-second chip is built for AI clusters. Terabits per second measure how much data the chip can handle. “Hyperscalers require networks that optimize latency, power and scalability simultaneously,” said Rishi Chugh, vice president and general manager of Marvell’s data-center switch unit. Marvell Technology, Inc.
But the trade looks crowded, and that’s a risk. A sharp drop Friday erased about $1.3 trillion in market cap from U.S.-listed chipmakers, with Marvell sliding 17%, according to Reuters. Ohsung Kwon, chief equity strategist at Wells Fargo, called the sector “way overbought.” Marvell’s latest filing notes it relies on a small group of big customers, is exposed to data-center sales, faces export controls, and warns AI spending on capex may not keep up. Reuters
Looking ahead, June 22 is the next key date. That’s when index demand and the AI trade meet the usual question about Marvell: how much of its future has the market priced in.
