Marvell shares surge 7% as AI hardware launch sparks analyst upgrades
Technology

Marvell shares surge 7% as AI hardware launch sparks analyst upgrades

Structera switch series designed to reduce power consumption and memory bottlenecks in data centers, with customer sampling set for third quarter

Marvell Technology shares rose 7% on Tuesday after the semiconductor company unveiled a new line of artificial intelligence hardware designed to reduce power consumption and simplify data center operations, prompting multiple Wall Street analysts to raise their price targets on the stock.

The new Structera series, announced at the Optical Fiber Communication Conference in Los Angeles, includes the Structera S 30260 CXL switch and the Structera S 60260 PCIe 6.0 switch. Both products address critical bottlenecks in AI workloads by improving memory sharing and reducing the complexity of connecting computing resources in large-scale data centers.

The Structera S 60260 represents the industry's first 260-lane PCIe 6.0 switch, offering nearly double the lane density of competing products, according to the company. This higher density enables data center operators to eliminate multiple smaller switches, resulting in reduced power consumption and lower latency. The switch can extend connectivity up to seven meters using active electrical cables and beyond that range with active optical cables.

The accompanying Structera S 30260 CXL switch supports CXL 3.0 standards and delivers aggregate bandwidth of up to 4TB/s. By enabling rack-level memory pooling, the device allows data centers to dynamically allocate memory resources across CPUs, GPUs, and other accelerators, addressing what Marvell describes as the "AI memory wall" that constrains performance in large-scale AI training operations.

"The Structera portfolio fundamentally changes the economics of AI data center infrastructure," the company stated in its press release. Both products are scheduled to begin customer sampling in the third quarter of 2026, with engineering test samples currently available at the OFC conference.

The product launch triggered a wave of analyst upgrades on March 6, with several firms raising their price targets significantly. J.P. Morgan increased its target to $135 from $130, maintaining an overweight rating. Craig-Hallum was more aggressive, lifting its target to $164 from $141, while B. Riley Securities raised its target to $135 from $130.

The consensus analyst target now stands at $120.28, implying approximately 31% upside from Marvell's current share price of $91.64. The stock has gained about 30% over the past year and is trading near its 52-week high of $102.70 reached earlier this month.

Marvell's fundamentals have strengthened alongside its AI-focused product strategy. The company reported trailing twelve-month revenue of $8.19 billion, with quarterly revenue growth accelerating to 22.1% year-over-year and earnings growth jumping 106.3% in the most recent quarter. The semiconductor maker currently trades at 28.6 times trailing earnings, with a forward multiple of 22.9.

Analyst sentiment remains broadly positive, with 33 analysts rating the stock a buy or strong buy compared to 11 hold ratings and no sell recommendations, according to market data. Institutional investors hold 83.6% of outstanding shares, indicating strong institutional confidence in the company's AI infrastructure positioning.

The Structera launch marks Marvell's latest effort to capture growing demand for specialized AI hardware. As data center operators race to build infrastructure capable of handling increasingly large AI models, the ability to reduce power consumption and improve memory efficiency has become a critical competitive advantage. Marvell's focus on connectivity and memory architecture places it in a distinct position within the AI semiconductor landscape, complementing rather than directly competing with the primary AI accelerator chips produced by companies like Nvidia and AMD.

The upcoming customer sampling period in the third quarter will serve as a key test for market adoption of the new technology. Successful deployment by major cloud providers or enterprise data center operators could validate Marvell's approach and provide further momentum for the stock, which has already outperformed the broader semiconductor index over the past year.