Jensen Huang points to Marvell as potential next “$1 trillion” AI chip winner, as Nvidia’s AI ecosystem deepens

Nvidia Chief Executive Jensen Huang has a new name in mind for the next major AI semiconductor winner. Huang said that Marvell Technology could become the next AI chip stock to reach a $1 trillion market capitalization, according to a Yahoo Finance report.

The comment is notable because it comes from the executive who has helped define the modern AI compute stack through Nvidia’s data center GPUs and networking products. Nvidia’s platform has become a centerpiece for training and running large AI models, and Huang’s willingness to spotlight a rival chip supplier underscores how the market for AI hardware is broadening across the stack.

Marvell Technology, a company known for switching, networking silicon, and custom silicon used in data centers, has increasingly positioned itself around the AI buildout. While the Yahoo Finance report focuses on the valuation-style prediction, it also indicates that investors are looking beyond any single supplier as AI clusters become more complex and demand dedicated components for data movement, acceleration, and interoperability.

For Nvidia, the practical implication is less about competition for the “next $1 trillion” and more about ensuring that its broader ecosystem remains central. As AI systems scale, workloads require not just raw compute, but also high-performance communication between accelerators and servers. Nvidia has built out networking and software layers intended to connect GPUs at scale, and those efforts can increase the importance of AI data-center platforms that include multiple vendors’ parts.

AI chip markets also tend to evolve from early winners to a wider set of specialized suppliers. Huang’s point, as reported, aligns with that pattern: power, performance, and integration matter across many components, so valuation leaps can occur for companies that capture a key piece of infrastructure, even if they are not the “primary” compute provider.

Still, details about the prediction were not disclosed in the Yahoo Finance summary beyond Huang’s characterization of Marvell’s potential. The report did not outline specific financial targets, product milestones, or timing for when Marvell might reach a $1 trillion market cap, and it did not provide additional context on which segment of Marvell’s business Huang was referring to.

Nvidia itself has been investing heavily to expand the AI data center footprint beyond just chips, including software capabilities and platform integration designed to support large-scale deployments. The broader message from Huang’s remarks is that the AI buildout is now large enough that multiple semiconductor companies could each become major platforms over time, rather than one company dominating all layers.

What to watch next is whether Marvell receives follow-on commentary from other industry leaders and whether investors begin to treat it less as a niche networking supplier and more as a full AI infrastructure contender. On Nvidia’s side, investors will likely continue to focus on how Nvidia’s performance and system-level offerings influence the adoption of multi-vendor AI server designs, including where Marvell’s components fit within those architectures.