Investment firms Evercore ISI and Melius Research have upgraded their outlooks for Marvell Technology, lifting price targets on the data infrastructure semiconductor provider.
The revisions reflect rising demand for AI infrastructure and an expected shift in AI workloads from training to inference by late 2026.
Evercore ISI raised its price target on Marvell Technology (NASDAQ:MRVL) to $155 from $133 on May 19, maintaining an Outperform rating.
The adjustment followed first-quarter AI channel checks that highlighted the anticipated transition to an inference-led market.
This shift intensifies the focus on cost-per-token, return on investment, and total cost of ownership. Hyperscalers are increasingly interested in internally developed ASICs and alternative accelerators.
Melius Research Also Boosts Target
On May 18, Melius Research analyst Ben Reitzes raised the firm's price target on Marvell to $220 from $140, keeping a Buy rating.
Reitzes noted that while nothing emerged as incrementally good from Trump's visit to China, the firm became incrementally positive on memory and AI semiconductor companies.
Melius subsequently raised long-term estimates and price targets for several Buy-rated "bottleneck stocks," including Micron, Sandisk, AMD, Intel, and Marvell, alongside a Hold-rated Qualcomm.
The firm believes semiconductor companies will capture more market capitalization growth than traditional software firms and non-semiconductor members of the Mag 7 over the long term.
Marvell Technology, Inc. designs, develops, and sells integrated circuits, supplying critical semiconductor solutions across the data center core and network edge.
The company boasts a five-year average revenue growth rate of 23.14 percent and secures a position among the 11 best long-term US stocks to buy, according to Detik Finance.
The price target upgrades come amid a broader trend of Wall Street analysts raising targets for AI-related semiconductor stocks.
The industry is closely watching the shift from training-focused AI workloads to inference, which could reshape demand for custom chips and accelerators.
