Home Artificial intelligence Not Nvidia. Not AMD. This Custom AI Chip Designer Could Outperform Both by 2030.
Artificial intelligence

Not Nvidia. Not AMD. This Custom AI Chip Designer Could Outperform Both by 2030.

Share


Nvidia (NVDA -0.80%) and Advanced Micro Devices (AMD -2.08%) are two of the biggest beneficiaries of the artificial intelligence (AI) revolution. As big tech invests hundreds of billions of dollars in data centers, demand for semiconductors continues to accelerate.

However, another chipmaker could deliver even greater returns over the next several years: Marvell Technology (MRVL -13.24%). Interestingly, Nvidia CEO Jensen Huang recognizes Marvell’s potential. During the Computex conference in June, Huang called Marvell the next potential trillion-dollar company. Nvidia has even invested $2 billion in Marvell, reinforcing its confidence in the company.

Now, Marvell is reportedly exploring an acquisition that could supercharge its position in the AI infrastructure market. Let’s dig into what makes Marvell unique and assess why the stock could be a compelling buy for long-term investors.

Marvell logo.

Image source: The Motley Fool.

What role does Marvell play in the AI chip stack?

Nvidia and AMD primarily design general-purpose AI accelerators, known as GPUs. Marvell specializes in designing custom application-specific integrated circuits (ASICs). These processors are tailored to specific workloads, potentially offering superior training and inference performance, energy efficiency, and economics compared with off-the-shelf chips.

This value proposition appeals to hyperscalers such as Amazon, Microsoft, Alphabet, and Meta Platforms. Rather than relying exclusively on products from Nvidia and AMD, big tech is also investing in custom silicon designed around their own AI models. Marvell is positioning itself as a partner in this transition.

Marvell’s opportunity extends beyond custom chips. Its portfolio also includes optical digital signal processors (DSPs), Ethernet switches, high-speed interconnects, and silicon photonics. These products help move data between larger clusters of GPUs and memory. As AI models evolve and use cases become more sophisticated, Marvell’s services should see increased demand.

Is Marvell acquiring Lumentum?

On Oct. 28, Reuters reported that Marvell was in advanced discussions to acquire Lumentum (LITE +0.61%), a manufacturer of optical components used in telecommunications networks and AI data centers. Wall Street initially reacted negatively, as Marvell stock declined more than 8% following reports of the discussions.

MRVL Marvell Technology

$218.34-$33.31 (-13.24%)

NASDAQ · As of 4:00 PM ET 10/29/26

The concerns are understandable. At the end of its fiscal second quarter, Marvell held approximately $3.9 billion in cash on its balance sheet. Lumentum boasts a market capitalization of $84 billion, meaning that an acquisition of this magnitude would require substantial borrowing, additional share issuance, or some combination of both by Marvell.

Even so, I see the strategic rationale here. One challenge facing AI data centers is moving information quickly and efficiently between large clusters of processors. Copper-based connections have limitations around bandwidth, distance, and power consumption. But optical interconnects offer an attractive alternative.

Lumentum specializes in lasers and optical components that support these systems. By integrating Lumentum’s technology with its networking and custom silicon portfolio, Marvell could offer customers a more comprehensive suite of AI infrastructure solutions.

On a deeper level, acquiring Lumentum would build on Marvell’s purchase of Celestial AI, a photonics company specializing in optical connectivity. In the long run, these investments could help Marvell become a critical supplier across multiple layers of the AI infrastructure stack.

Can Marvell become a trillion-dollar company?

Huang’s prediction might sound ambitious at first glance, but Marvell’s long-term financial projections help explain his enthusiasm. During its investor day earlier this month, Marvell’s management outlined a path to reach between $70 billion and $90 billion in annual revenue by fiscal 2031, which largely corresponds to calendar 2030. At the midpoint, this implies $80 billion in annual sales.

For perspective, Marvell generated $8.2 billion in revenue during fiscal 2026. Reaching management’s midpoint forecast would represent a tenfold increase over just five years!

The company believes this growth could support roughly $30 in adjusted earnings per share (EPS) by fiscal 2031. If Marvell achieves this target and is able to command a price-to-earnings (P/E) multiple of 30, its stock would trade around $900 per share. That translates into an implied market capitalization of roughly $810 billion, or roughly 309% upside from today’s price action.

This analysis is simply here to illustrate that Marvell offers compelling long-term prospects, regardless of whether it joins the trillion-dollar club. While Nvidia and AMD will remain important players in AI computing, Marvell’s assembly of custom silicon, advanced networking, and optical connectivity could allow it to capture an increasingly valuable portion of the data center market.

For investors who can stomach some volatility as the AI infrastructure era enters a new phase, I see Marvell as one of the best AI semiconductor stocks to watch.



Source link

Leave a comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Articles
Artificial intelligence

Lagos non-profit, engineering.fit( ) trains young Nigerian engineers in machine learning research

By Emmanuel Okogba A Lagos-based non-profit organisation, engineering.fit( ), is providing free...

Artificial intelligence

Microsoft, Google report booming AI revenue as the tech reshapes business

This story was originally published on CIO Dive. To receive daily news...

Artificial intelligence

Google Eats Its Own Tail (and Profits) With AI Summaries as Stock Price Plunges

Big tech firms have poured an infinite amount of money into artificial...

Artificial intelligence

UK Government AI investment 2026: Free courses, workforce plans explained

Partner contentWhy colleges are rewriting IT budgets this term Skills Bootcamps now...