Jensen Huang Just Said This AI Chip Stock Could Be the Next $1 Trillion Company (Hint: It's Not AMD or Sandisk)

Jensen Huang called Marvell "the next trillion-dollar company," helping drive the stock up more than 30% intraday and reinforcing investor optimism around its AI-linked growth story. Nvidia also disclosed a $2 billion strategic investment and expanded partnership with Marvell across NVLink Fusion, optical interconnects, and silicon photonics. Marvell's market cap is still only about $275 billion, but fiscal 2026 revenue rose 42% to just under $8.2 billion and Q1 fiscal 2027 revenue reached $2.4 billion, with management guiding data center growth of 40% and interconnect growth of 50% this year.

Analysis

This is less about a single endorsement and more about Nvidia signaling a supply-chain architecture it intends to own: compute, rack interconnect, and now the plumbing around the rack. If that stack keeps consolidating, the incremental value accrues to the “boring” infrastructure layer, where switching costs rise and design wins compound over multiple procurement cycles. That is bullish for the partner network, but it also raises the bar for every adjacent vendor not already embedded in Nvidia-led reference designs.

The immediate winner is MRVL, but the second-order effect is compression of differentiation across the broader AI networking cohort. A premium rerating can persist for several quarters if Marvell converts the endorsement into visible backlog and gross margin stability, but the current setup is already pricing in a lot of execution. The bigger tell will be whether hyperscalers treat this as a single-vendor stack optimization or use it to extract price concessions from other interconnect and custom silicon suppliers.

The main risk is not that the story breaks overnight; it is that expectations outrun the cadence of fundamental inflection. At this valuation, any delay in monetizing the partnership, any customer concentration hiccup, or any sign that insider selling is front-running a slower ramp could trigger a sharp multiple reset even if revenue keeps growing. Over the next 3-6 months, the stock is likely more sensitive to guidance revisions and design-win commentary than to headline partnership news.

Contrarianly, the market may be underestimating how much of this narrative is already a function of Nvidia’s halo rather than Marvell-specific fundamentals. If investors begin to view MRVL as a leveraged proxy for Nvidia capex rather than a standalone compounder, the beta can work both ways. That makes the setup attractive tactically, but not necessarily as a clean long-term compounder at any price.

Key Decisions for Investors

• Long MRVL on pullbacks only, with a 4-8 week horizon; prefer entry after post-gap consolidation rather than chasing the endorsement spike. Upside is continued multiple expansion if management confirms partnership-driven backlog, but downside is large if the stock loses momentum at >70x forward earnings.
• Pair trade: long NVDA / short MRVL for 1-3 months if you want exposure to AI infrastructure with less execution risk. NVDA monetizes the ecosystem regardless, while MRVL is already priced for near-flawless follow-through.
• Use MRVL call spreads rather than outright calls for the next earnings cycle. The setup favors volatility around guidance, but implied expectations are already elevated, so defined-risk upside captures more of the move if the market extends the rerate.
• Watch for weakness in competing AI interconnect and custom-silicon names over the next quarter; any sign of budget reallocation toward Nvidia-aligned architecture is a relative negative for smaller networking suppliers that lack ecosystem leverage.
• If MRVL adds another 15-20% without a new fundamental catalyst, begin scaling out. The risk/reward becomes materially less attractive once the market fully prices the endorsement and not just the earnings path.