Broadcom jumps 6%, Marvell soars 17% on twin AI catalysts

Two AI chip beneficiaries surged in premarket trading Tuesday as Alphabet's $80 billion capital raise and Nvidia CEO Jensen Huang's trillion-dollar endorsement converged on the custom-chip sector.

Broadcom Inc. rose 6% and Marvell Technology Inc. jumped 22% in premarket trading Tuesday after Alphabet Inc. announced an $80 billion equity raise for AI infrastructure and Nvidia CEO Jensen Huang predicted Marvell could reach a $1 trillion valuation.

"You're going to be the next trillion-dollar company," Huang told Marvell CEO Matt Murphy during a public appearance at the Computex conference in Taiwan, according to a transcript of the exchange.

Broadcom shares traded near $486 after closing Monday at $459.97, extending a year-to-date gain of 33%. Marvell stock surged to about $257, pushing its 2026 return past 158%. Alphabet's equity issuance — which includes a $10 billion private placement to Berkshire Hathaway Inc. — will fund investments in AI compute infrastructure, directly benefiting Broadcom as Google's custom TPU design partner under an agreement extending through 2031.

The twin catalysts highlight the scale of capital flowing into AI chipmaking. Broadcom's AI semiconductor revenue jumped 106% year over year to $8.4 billion in its fiscal first quarter, with CEO Hock Tan guiding for $10.7 billion in the current period. For Marvell, Huang's public endorsement could accelerate a re-rating in the merchant custom-silicon market, where the company competes with Broadcom for cloud customers designing their own chips.

Alphabet's $80 Billion Bet on Custom Silicon

Alphabet's decision to raise $80 billion through an equity offering — the largest single capital raise by a US technology company — signals that the search giant expects AI infrastructure demand to remain elevated for years. The proceeds will fund data center construction, GPU procurement, and custom TPU development, the latter of which flows directly to Broadcom's semiconductor business.

The market reaction for Alphabet shares was more muted: GOOGL fell 2.5% in premarket trading as investors weighed dilution from the new shares. Berkshire Hathaway's $10 billion participation provided a vote of confidence from Warren Buffett's conglomerate, which has historically avoided large technology equity issuances.

Huang's Endorsement Reshapes the Custom-Chip Narrative

Jensen Huang's prediction that Marvell could join the trillion-dollar club is notable not just for its ambition but for its source. Nvidia, the world's most valuable company at roughly $4.5 trillion, has little to gain — and potentially something to lose — from promoting a merchant custom-chip rival. Marvell designs application-specific integrated circuits (ASICs) for cloud customers, a market that competes with Nvidia's general-purpose GPU business for AI workloads.

Marvell's 158% year-to-date surge already reflects aggressive rerating tied to its custom-silicon pipeline. The company has secured design wins with multiple hyperscale cloud providers, though it has not disclosed the full scope of its contract pipeline. Huang's endorsement, delivered on stage at Computex, provides the kind of third-party validation that can attract institutional investors who have been waiting for confirmation of the thesis.

What This Means for Investors

The convergence of Alphabet's capital commitment and Huang's endorsement creates a powerful narrative for the custom-chip sector. Broadcom trades at roughly 28x forward earnings, reflecting its established position as Google's TPU partner and its diversified software business. Marvell, by contrast, trades at a higher multiple, reflecting the growth optionality embedded in its custom-silicon pipeline.

Both companies face execution risk: Broadcom must deliver on its $10.7 billion AI semiconductor guidance while managing supply chain constraints at TSMC, its primary foundry partner. Marvell must convert its design wins into production revenue at a pace that justifies its valuation. For now, the market is betting that the AI infrastructure buildout has years of runway ahead.

This article is for informational purposes only and does not constitute investment advice.