Marvell Technology: The Next AI Chip Giant by 2030?

Nvidia and AMD are currently leading the charge in the artificial intelligence (AI) chip market, benefiting immensely from major tech investments in data centers. However, there's another player in the semiconductor space that might just outshine them: Marvell Technology.

Marvell's Growing Significance in AI

At a recent Computex event, Nvidia’s CEO, Jensen Huang, forecasted Marvell's potential to become a trillion-dollar enterprise, highlighting the confidence Nvidia has placed in the company through a $2 billion investment. But what sets Marvell apart in the quickly evolving landscape of AI chip technology?

While Nvidia and AMD focus on general-purpose graphics processing units (GPUs), Marvell specializes in creating custom application-specific integrated circuits (ASICs). These chips are designed for particular applications, which provides better efficiency in energy usage and performance. This specialization makes Marvell an attractive option for tech giants like Amazon, Microsoft, Alphabet, and Meta, which are looking to optimize their AI capabilities.

Portfolio That Goes Beyond Custom Chips

Marvell's innovations aren't limited to just custom silicon. The company also develops optical digital signal processors (DSPs), Ethernet switches, and high-speed interconnects to facilitate data movement between extensive arrays of GPUs and memory systems. As the AI landscape evolves, the demand for Marvell’s advanced products is expected to grow.

Potential Acquisition of Lumentum

Newly emerging signals point towards Marvell considering an acquisition of Lumentum, a producer of optical components critical for telecommunications and AI data centers. Although the stock price dropped by over 8% in response to these discussions, this acquisition could be pivotal.

Marvell currently boasts around $3.9 billion in cash, while Lumentum holds a market cap of $84 billion. Funding such an acquisition would likely involve borrowing or issuing new shares. Yet, from a strategic viewpoint, acquiring Lumentum could enhance Marvell’s capabilities in facilitating high-speed data transfer within AI data centers.

Copper-based connections face issues regarding bandwidth and power efficiency, whereas optical interconnects present a more effective solution. By merging Lumentum's optical technology with its existing portfolio, Marvell could strengthen its position as a comprehensive provider for AI infrastructure.

Long-Term Growth Potential

Huang's enthusiastic outlook on Marvell may seem ambitious, but the company’s projections for its financial future support his optimism. At a recent investor meeting, Marvell’s management outlined a forecast predicting annual revenues between $70 billion and $90 billion by fiscal year 2031, which coincides with roughly calendar year 2030. If achieved, that means a staggering increase from the $8.2 billion revenue generated in fiscal 2026.

The company anticipates that this growth trajectory could lead to adjusted earnings per share of around $30 by fiscal 2031. If Marvell captures a price-to-earnings multiple of 30, the stock could reach approximately $900 per share, suggesting a potential market capitalization swelling to around $810 billion. This projection indicates a sizable upside from current figures.

While the prospects of joining the trillion-dollar club remain uncertain, Marvell possesses encouraging long-term potential. The integration of custom chips, advanced networking, and optical technologies positions Marvell to claim an increasingly valuable share of the data center market.

For investors ready to navigate the fluctuations brought on by the AI infrastructure boom, Marvell stands out as a promising stock to monitor.