Jensen Huang's endorsement of Marvell Technology at Computex 2026 on June 1 gave investors a sharp reminder that the AI infrastructure trade extends well beyond Nvidia itself. Huang called Marvell Technology a potential “next trillion-dollar company” during an onstage appearance with Marvell CEO Matthew Murphy in Taipei.

The comment sent Marvell custom AI silicon into the spotlight and added roughly 17% overnight before extending gains during Tuesday’s session. For investors, the real question is not whether Huang was being polite to a partner. The question is whether Marvell Technology, Inc. (NASDAQ:MRVL) has the financial trajectory to back up that claim.

The Gap Between $180 Billion and $1 Trillion

Marvell currently carries a market capitalization of approximately $192 billion. Reaching $1 trillion from that base implies roughly a five-fold increase. Admittedly, that sounds steep. However, context matters here. Marvell posted record fiscal 2026 revenue of $8.195 billion, a 42% year-over-year increase driven by robust AI demand. In Q1 fiscal 2027, the company reported record revenue of $2.418 billion, up 28% year-over-year, with record operating cash flow of $638.8 million.

Furthermore, Marvell guided Q2 fiscal 2027 revenue to $2.7 billion at the midpoint, representing 35% year-over-year growth, and significantly raised its revenue outlook for both fiscal 2027 and fiscal 2028. Sustained growth at that rate compresses the timeline to trillion-dollar territory faster than many investors assume. Marvell custom AI silicon is therefore not a speculative story. It is a revenue story with visible momentum.

Why the Nvidia Partnership Is Structurally Important

On March 31, 2026, Nvidia Corporation (NASDAQ:NVDA) and Marvell announced a strategic partnership connecting Marvell to the Nvidia AI factory and AI-RAN ecosystem through the NVLink Fusion platform. Nvidia also invested $2 billion in Marvell as part of that arrangement. This is not a routine supplier relationship. NVLink Fusion allows hyperscalers to build semi-custom AI infrastructure using Nvidia's ecosystem. Under the partnership, Marvell provides custom XPUs and NVLink Fusion-compatible scale-up networking. Nvidia contributes Vera CPUs, ConnectX NICs, Bluefield DPUs, and Spectrum-X switches.

The arrangement gives Marvell deep integration into the dominant AI compute stack. It also positions Marvell custom AI silicon as a complement to Nvidia's GPUs rather than a competitor. That is a strategically safer place to operate.

Hyperscaler Wins Validate the Custom Silicon Roadmap

Marvell's data center segment generated $1.832 billion in Q1 fiscal 2027, representing 76% of total revenue, up 27% year-over-year. That concentration reflects how central custom silicon has become to Marvell's business. The company designs chips for Amazon's Trainium and Inferentia processors and Microsoft's Maia AI accelerator. Additionally, Google is reportedly in talks with Marvell to develop two new AI chips: a memory processing unit and an inference-optimized TPU.

Those discussions would add Marvell as a third design partner alongside Broadcom and MediaTek in Google's custom chip supply chain. These wins reduce single-customer risk meaningfully. Three anchor hyperscalers provide revenue visibility that single-customer ASIC stories historically lacked. Broadcom Inc. (NASDAQ:AVGO) and Marvell together control an estimated 95% of the custom AI ASIC co-design market, a duopoly that is difficult for new entrants to break quickly.

Competitive Positioning Against Broadcom

Marvell custom AI silicon sits in a well-defined competitive frame. Broadcom is widely cited as holding roughly 70% of the custom AI accelerator market, anchored by long-running programs with Alphabet Inc. (GOOGL) for TPU silicon and Meta Platforms Inc. (META) for its MTIA chips. Marvell holds a smaller but growing share, estimated at approximately 25% of the market by 2027 according to Counterpoint Research. The margin profiles of both companies also differ. Broadcom runs at roughly 60% non-GAAP operating margin, partly reflecting software revenue from VMware. Marvell sits in the high 20s to low 30s, with room to expand as AI mix grows.

That margin gap means Marvell has a larger re-rating opportunity ahead as custom silicon scales. Marvell also moved into optical interconnects through its acquisitions of Celestial AI and XConn Technologies, both completed in February 2026. Those deals target silicon photonics, which uses light rather than electrical signals to move data between chips. Silicon photonics is considered a key efficiency lever for next-generation AI data centers.

What Matters to Investors

Several near-term catalysts give the Marvell custom AI silicon thesis near-term visibility. Marvell forecast scale-out switching revenue to exceed $600 million in fiscal 2027, roughly doubling from fiscal 2026, with a path to more than $1 billion in annualized revenue in fiscal 2028. TrendForce projects custom chip sales will increase 45% in 2026, compared with 16% growth in GPU shipments.

That divergence means custom silicon is outpacing the broader chip market. Still, investors should note key risks alongside the opportunity. Marvell's revenue is heavily concentrated in the data center segment. Customer concentration remains a concern, as a delay in one hyperscaler's capital spending cycle can move Marvell's quarterly results meaningfully. The stock also trades at a significant premium on a price-to-earnings basis following its year-to-date rally of over 185%. Margin expansion is still a work in progress, and the Celestial AI acquisition adds integration execution risk.

Bottom Line

Huang's trillion-dollar comment is a headline. The underlying financial architecture of Marvell makes the claim worth taking seriously. Design wins in fiscal 2026 hit an all-time record, and Marvell management expects those wins to continue fueling future growth. The company sits at a structural intersection of custom silicon, optical interconnects, and AI networking. Each of those segments is growing faster than the broader semiconductor market. Investors watching the AI infrastructure trade who want exposure outside of Nvidia's direct orbit should keep Marvell on their radar. The trillion-dollar road is long. The quarterly data suggests Marvell is traveling it at speed.

Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.