Marvell on the Eve of Earnings: Wall Street Collectively Raises Price Targets, Can Nvidia and AMD’s Double Endorsement Deliver on the AI Narrative?

The evolution of the generative AI boom is driving a shift in computing power demand from training-centric to inference-centric, a trend creating vast market space for AI ASIC chips. Compared to general-purpose GPUs, AI ASICs customized for specific workloads possess inherent advantages in key metrics such as cost per token, power consumption, and memory bandwidth utilization, making them a critical choice for cloud service providers to reduce AI deployment costs.

As a pioneer in the AI ASIC field, Marvell Technology has deeply positioned itself within this track. The company is not only the core partner for Amazon's ( AMZN) AWS Trainium series AI chips, but also participated in Google's ( GOOGL) Axion ARM CPU design, Microsoft's ( MSFT) Maia AI accelerator development, and Meta's new data processing unit customization projects.

Citi significantly raised the stock's price target from $118 to $215 on May 21, reiterating a "Buy" rating. The core logic lies in the sustained strong demand for Amazon's Trainium2 ASIC, prompting analysts to upwardly revise earnings forecasts and maintain a robustly optimistic outlook for the upcoming quarterly results.

Following closely, Stifel raised the price target from $140 to $210 on May 22, also maintaining a "Buy" rating. The firm highlighted Marvell's collaborative relationship with NVIDIA, which, combined with upward revisions in capital expenditure expectations from hyperscale cloud providers, further solidifies the medium-to-long-term growth curve of its data center business; the company is expected to deliver financial results that exceed market consensus.

Wells Fargo previously also raised the price target from $135 to $195, assigning an "Overweight" rating. The bank noted that based on expected 2027 earnings, the current valuation corresponds to a price-to-earnings ratio exceeding 30x, which creates a higher bar for expectations regarding short-term earnings releases.

Evercore ISI raised Marvell's price target from $133 to $155 on May 19, maintaining an "Outperform" rating. Through first-quarter AI channel research, the firm found that AI workloads are accelerating their shift from training-centric to inference-centric, a trend expected to solidify by the end of 2026. The cost-efficiency considerations brought by this transition have significantly increased demand from hyperscale customers for in-house ASICs and other alternative accelerators, which perfectly aligns with Marvell's core business layout.

On the same day, Melius Research significantly raised Marvell's price target from $140 to $220, maintaining a "Buy" rating.

Beyond AI chips, Marvell Technology's leadership in high-speed data center interconnects is also highly recognized by the market. When data center connection distances exceed 10 meters, copper cables can no longer meet bandwidth demands, making optical interconnects the inevitable choice.

Leveraging its technological advantages in optical DSP and optical interconnect products, Marvell has become a core supplier in the global optical interconnect market.

More importantly, Marvell's silicon photonics technology extends optical interconnects from inter-rack to intra-rack, intra-system, and even within chip packaging, providing more efficient transmission solutions for AI data centers.

Recently, the company further strengthened its technological footprint in the silicon photonics space through the acquisitions of Celestial and Polariton Technologies, a Swiss high-speed optical chip firm.

Management's previously provided performance framework offered clear market guidance, projecting that data center revenue will grow by approximately 40% in fiscal 2027, with interconnect revenue expected to increase by over 50% year-over-year—significantly higher than the prior 30% estimate.

Institutional feedback has also been positive; Oppenheimer raised its price target for Marvell from $170 to $200 on May 20, maintaining an "Outperform" rating while highlighting the company's substantial market share and growth potential in the wired networking market.

The key to the upcoming first-quarter earnings report lies in whether management can demonstrate to investors that businesses such as optical interconnects, silicon photonics, CXL, and PCIe switching are becoming genuine drivers of data center growth, rather than just a long-term AI conceptual narrative.

Year-to-date, Marvell Technology's stock price has surged by as much as 120%, repeatedly hitting all-time highs, with market confidence fully reflected in its price action.

Meanwhile, the positions held by two chip giants, AMD ( AMD) and Nvidia ( NVDA ), have added significant weight to this confidence. AMD's recently disclosed stake has triggered market speculation regarding potential strategic integration, while Nvidia's $2 billion strategic investment provides a dual endorsement for the company from both technical and ecosystem perspectives.

On May 12, AMD disclosed its position in Marvell in a regulatory filing, a move that quickly sparked heated market discussion. The following day, Marvell's shares surged 8.18% to close at $177.95, a record high.

The market immediately linked this move to a potential strategic acquisition—AMD has a deep presence in data center CPUs and GPUs but has shortcomings in AI interconnect technology, a field where Marvell is an absolute leader.

Although neither party has issued an official statement regarding a merger or acquisition, AMD's stake itself sends a clear signal that the chip giant, with a market capitalization of approximately $400 billion, is paying close attention to Marvell's strategic value.

Compared to AMD's subtle positioning, Nvidia's move was more direct.

On March 31, Nvidia and Marvell announced a strategic partnership where the former integrated Marvell into its AI Factory and AI-RAN ecosystem through the NVLink Fusion platform and completed a $2 billion strategic investment.

For Marvell, the value of this deal goes far beyond the capital itself. Technical integration places Marvell at the core intersection of both GPU and ASIC ecosystems; NVLink Fusion allows customers to flexibly deploy Marvell's custom chips and networking products on top of Nvidia's architecture to build heterogeneous AI infrastructure. Simultaneously, the $2 billion strategic investment represents Nvidia's highest recognition of Marvell's technological standing.

This content was translated using AI and reviewed for clarity. It is for informational purposes only.