Written by Cassian Vance
When the most powerful figure in the global technology sector speaks, the market listens. But when Nvidia CEO Jensen Huang explicitly points to a partner company and declares it “the next trillion-dollar company,” the market explodes.
On June 2, 2026, at the Computex conference in Taipei, Huang shared the stage with Marvell Technology (MRVL) CEO Matthew Murphy. During their appearance, Huang delivered a sweeping endorsement of Marvell’s role in the artificial intelligence infrastructure buildout. The reaction on Wall Street was historic: Marvell shares surged 33% in a single session, closing at a 52-week high of $291.20 and adding over $62 billion in market capitalization.
The endorsement validates a thesis that institutional investors have been building for months: as AI workloads scale, the bottleneck is no longer just compute power, but connectivity.
The Connectivity Bottleneck
The core of Huang’s argument centers on how data centers are evolving. “When you take a computing problem, and you disaggregate it into a lot of parts, and you distribute it across the entire data center, what’s necessary is connectivity,” Huang stated. “That’s the reason why Matt’s doing so well. That’s the reason why Marvell is so essential.”
As generative AI models become exponentially larger, they cannot be trained or run on a single chip, or even a single server rack. The workloads must be distributed across thousands of GPUs working in tandem. Marvell specializes in the optical interconnects, digital signal processors (DSPs), and custom silicon required to move massive amounts of data between these chips without latency.
This architectural shift is driving unprecedented financial results. In late May, Marvell reported a record fiscal Q1 2027 revenue of $2.418 billion, representing a 28% year-over-year increase. The Data Center segment alone accounted for $1.833 billion of that total.
| Metric | Q1 FY2027 Result | YoY Growth |
| Total Revenue | $2.418 Billion | +28% |
| Data Center Revenue | $1.833 Billion | N/A |
| Non-GAAP EPS | $0.80 | Beat Estimates |
Management’s forward guidance is equally aggressive. CEO Matt Murphy cited “exceptional AI-related bookings” and raised the full-year fiscal 2027 outlook to approximately $11.5 billion. More notably, Marvell expects its custom chip revenue alone to exceed $10 billion by fiscal 2029, with total company revenue targeting $15 billion by fiscal 2028.
The Nvidia Strategic Alliance
Huang’s Computex endorsement was not merely rhetorical; it is backed by capital and deep technical integration. In March 2026, Nvidia committed a $2 billion strategic investment into Marvell.
The partnership centers on NVLink Fusion, a platform that allows Marvell’s custom Application-Specific Integrated Circuits (ASICs) and optical engines to slot seamlessly into Nvidia’s proprietary rack architectures. While hyperscalers like Alphabet, Meta, and Amazon are designing their own custom chips to reduce reliance on expensive general-purpose GPUs, they still need those custom chips to communicate within Nvidia-dominated data centers. Marvell provides the crucial translation and connectivity layer.
Recent acquisitions of Celestial AI (photonics) and XConn Technologies (chiplets) have further widened Marvell’s moat, giving the company hard-to-replicate intellectual property in 5nm SerDes and CMOS optical engines.
Valuation and Execution Risk
Following the 33% single-day surge, Marvell’s market capitalization stands at roughly $250 billion. To fulfill Huang’s prophecy and reach the $1 trillion mark, the stock would need to quadruple from current levels.
At $291 per share, Marvell trades at a trailing price-to-earnings ratio near 100x. The market has aggressively priced in years of flawless execution. If hyperscalers begin to bring more silicon design in-house, or if the broader AI capital expenditure cycle cools, Marvell’s premium multiple would contract violently.
Furthermore, Marvell faces intense competition from Broadcom (AVGO), the incumbent giant in custom silicon. Broadcom is roughly five times Marvell’s size by revenue and boasts deep, entrenched relationships with Google and Meta. While Marvell is the agile challenger tightly aligned with Nvidia, Broadcom remains the dominant force in the space.
The Verdict: Sizing the Catalyst
Jensen Huang’s endorsement has permanently altered Marvell’s trajectory, transforming it from a derivative AI play into a foundational infrastructure asset. The custom silicon and optical connectivity thesis is fully intact. However, chasing a 33% single-day gap-up carries significant near-term risk.
Marvell Technology (MRVL) Verdict: BUY on Pullbacks
- Current Price:** ~$291.20
- Price Target:** $330.00
- Catalyst:** Accelerating optical interconnect deployments and expanding NVLink Fusion integration with Nvidia.
The Infrastructure Ecosystem
The Computex announcements and subsequent market reaction have recalibrated the AI hardware landscape.
Broadcom (AVGO) Verdict: BUY
Broadcom shares rose 6% on June 2 following Alphabet’s announcement of an $80 billion stock sale to fund AI infrastructure. Broadcom remains the undisputed leader in merchant custom silicon and reports its highly anticipated Q2 earnings today (June 3). It is the safer, more diversified play in the custom ASIC space.
Astera Labs (ALAB) Verdict: BUY
As a provider of connectivity silicon for PCIe and CXL fabrics, Astera Labs benefits from the exact same data center bottleneck that Marvell is solving. It offers adjacent exposure to the AI connectivity buildout without competing directly with Marvell on ASIC design.
Hewlett Packard Enterprise (HPE) Verdict: HOLD
HPE surged 19% on June 2 following a strong earnings beat driven by AI server demand. While the quarter was impressive, HPE lacks the proprietary silicon moats enjoyed by Marvell and Broadcom, making it more vulnerable to hardware commoditization.
__________________________________________________
Disclaimer: This article is for informational purposes only and does not constitute investment advice. The opinions expressed are those of the author and do not reflect the views of Equities Orbis or its affiliates. Always conduct your own research before making investment decisions.
