ASIC and optical interconnect underpin a high-speed growth landscape! BofA strongly backs the surging Marvell (MRVL.US), proclaiming there is still 55% upside ahead.
Bank of America maintains its $365 price target for Marvell, emphasizing its ability to expand revenue share per AI system through custom AI accelerators (i.e., AI ASIC/XPU) and accompanying optical interconnect chips, driven by massive AI agent inference demand; relative to the September 11 closing price of $236.10, this target implies approximately 54.6% potential upside.
Title context: ASIC and optical interconnect underpin a high-speed growth landscape! BofA strongly backs the surging Marvell (MRVL.US), proclaiming there is still 55% upside ahead.
Text:
On September 11, the U.S. reported that headline and core CPI for August rose 0.4% and 0.3% month over month, respectively, both faster than in July, with core CPI exceeding expectations on a month-over-month basis. Yet that same day, the S&P 500 and Nasdaq Composite still rose 0.86% and 0.96%, respectively, as falling oil prices and strong earnings from AI-related technology companies continued to provide bullish support, highlighting that the market is still weighing inflationary pressures under the GEO Group Inc political conflict against earnings growth trends. Wall Street's bullish logic centered on Marvell Technology, Inc. (MRVL.US), which focuses on AI ASIC and data center optical interconnect chips, precisely reflects the latterthat is, "an earnings trajectory dominated by the AI compute theme against all negative logic."
Wall Street financial giant Bank of America Corp maintained its $365 price target on Marvell, valuing its ability to expand revenue share per AI system through custom AI accelerators (i.e., AI ASIC) and accompanying optical interconnect chips built around massive AI agent inference demand; relative to the September 11 closing price of $236.10, that target implies about 55% potential upside. However, the $40 billion to $45 billion related revenue opportunity in 2030 is a scenario estimate, and the extent to which it materializes depends on customer projects, product share, and profitability.
The larger the AI compute demand, the more the value of data center AI chips and optical interconnect expands
From an engineering and business model perspective, the core of Marvell's benefit from AI expansion is "custom AI accelerated computing + system-supporting optical interconnect/optical communications." ASICs designed for specific workloads can improve energy efficiency and reduce large-scale deployment costs by optimizing compute units, memory access, and data flow; at the same time, more accelerators urgently need fully supporting networking, memory interface chips, and high-speed data center connectivity chips based on optical technology.
Marvell calls this supporting business "XPU attach chips," covering PCIe retimers, coprocessors, CXL controllers, high-speed optical interconnect chips, and key components for co-packaged optics (CPO). According to Bank of America Corp's latest estimates, one to two attach chips priced at $500 to $1,500 per custom processor can give the company multiple revenue increments as customers scale. The complexity of high-speed analog circuits, interface intellectual property, and system validation also creates barriers for customers to replicate these products on their own.
Marvell's second growth curvethat is, in addition to building AI ASIC chips with cloud computing giants such as Amazon.com, Inc., its "optical interconnect chip business"has an underlying DRIVE in optical interconnect demand: the effective compute power of AI clusters increasingly depends on cross-node communication efficiency. Expert parallelism in mixture-of-experts models, distributed inference, and systems using prefill and decode separation may all increase data exchange; network congestion can leave expensive accelerators waiting for data and reduce utilization across the entire cluster.
As transmission rates and distances increase, the loss, signal compensation, and power consumption pressures of copper connections intensify, pushing more links to adopt optical connections. Marvell participates through optical communications digital signal processors, high-speed serial interfaces, drivers, and related interconnect products. The company's already mass-produced Ara platform and expanded 1.6T product portfolio are concrete products capturing this upgrade.
Lumentum's demand outlook provides evidence from another side of this supply chain. Management judgments recorded in Citi meeting notes include: shipments of related optical module devices (i.e., optical modules used for tightly coupled interconnect with AI accelerators) are expected to roughly double from 2026 to 2027, with shortages of EML and CW lasers expected to persist through 2027, while improved visibility in OCS and NPO businesses drives an upward revision of the fiscal 2028 earnings per share target to $40.
The two companies benefit in different parts of the chain: Marvell focuses on providing semiconductors that process, transmit, and connect data, while Lumentum provides laser light sources, optical components, and optical switching systems. Therefore, the growth logic for AI chips and optical interconnect still has a basis for continuation, but sustained stock price gains require continuous delivery on orders, capacity certification, gross margins, and cash flow.
Bank of America Corp sends a strong bullish signal to Marvell shareholders
According to Seeking Alpha statistics, this AI chip giant's stock price has risen more than 160% over the past six months, with demand for its data center chips fueling market optimism. According to a report shared with TheStreet, Bank of America Corp analyst Vivek Arya, after meeting with CEO Matt Murphy and CFO Dan Durn, believes there is ample reason to remain firmly bullish on the stock.
Wall Street's overall view of Marvell is bullish, but price targets diverge. As of September 11, the S&P Global, Inc. consensus rating from 45 analysts shown by StockAnalysis was "Strong Buy," with an average one-year price target of $284.64, corresponding to about 20.6% upside from the $236.10 closing price, while the negative price target implies a decline of about 20%. The highest target in that sample was $400, corresponding to about 69.4% potential upside; public rating records show that KeyBanc analyst John Vinh maintained an "Overweight" rating and a $400 price target on August 28. BofA's $365 target is also clearly above the market average. Wall Street's common bullish direction is expansion in the AI data center business, with disagreement centered on the pace of growth realization and the valuation multiple the market is willing to assign.
As of September 11, the stock had risen nearly 12%, while shares of large semiconductor peers such as NVIDIA Corporation (NVDA.US) had fallen more than 5% over the past week. Specifically, Marvell participates in building the hardware that supports AI operations. The company designs custom processors, as well as high-speed optical interconnect chips that transmit data between processors and enable massive computing systems to work together efficiently.
The company's expanded cooperation with Alphabet Inc. Class C is particularly noteworthy, covering chips that support the search giant's AI infrastructure. In addition, the agreement also gives Alphabet Inc. Class C the right to purchase Marvell stock at a specific price, depending on future procurement.
But Arya's latest view, published after lunch with management, already goes beyond any single customer. BofA emphasizes that Marvell's expanding product portfolio may further deepen its role in AI systems. With the October 6 analyst day approaching, investors may soon get a clearer view of just how large this opportunity is.
Bank of America Corp maintained its "Buy" rating and high $365 price target on Marvell stock, implying nearly 55% upside compared with the stock's September 11 closing price of about $236.10.
The analyst team led by Vivek Arya bases its confidence on Marvell's potential to sell more technology products into each AI system.
Arya and his team believe Marvell's biggest opportunity lies in the attach chips that connect processors, manage memory, and transmit data. Compared with the processors themselves, this business is much harder for customers to replicate.
Marvell has already shipped these products to the four major U.S. hyperscale cloud providers, and each of its custom processors requires one to two attach chips priced at $500 to $1,500. This creates a new opportunity for Marvell to increase revenue from these customers as they expand compute capacity.
BofA estimates that by 2030, this attach chip market alone could exceed $60 billion to $65 billion. Assuming a 40% to 50% market share, Marvell's annual sales opportunity could reach $30 billion, while management's outlook for 2028 is $3 billion to more than $4 billion.
In addition, custom processors provide another important growth driver. BofA believes that by 2030, Marvell's potential sales in this area could reach $15 billion, bringing the combined opportunity to $40 billion to $45 billion. However, it should be noted that these are modeled opportunities, not confirmed orders.
The earnings potential also explains this optimism. BofA analyst Arya believes Marvell's earnings per share potential in calendar 2028 is close to $14, above the baseline level of $11; for every additional $1 billion in sales, earnings per share could increase by 30 to 35 cents.
His price target uses a 33x valuation on baseline earnings excluding stock-based compensation expense. Therefore, the bull case depends heavily on strong growth and on investors continuing to award Marvell a premium valuation as it executes its plan.
For reference, according to Seeking Alpha data, the stock's non-GAAP price-to-earnings ratio over the past five years was 42x, so the 33x baseline valuation requirement is not that demanding.
What factors could interrupt Marvell's rally?
The biggest risk at the stock price level for Marvell is the gap between its enormous market opportunity and actually realized earnings. The massive $40 billion to $45 billion sales scenario envisioned by BofA depends heavily on market expansion, customer spending, and a substantial stock price increase. These sales revenues have yet to be realized.
Execution on customer projects can be described as the first test. BofA noted uncertainty in next-generation chip projects at Amazon.com, Inc. and Microsoft Corporation. Product launch delays could push sales revenue out, while R&D costs would continue to be incurred, weakening the earnings growth investors expect.
Competition is likely to further intensify this risk. Share leader in AI ASIC Broadcom Inc. (AVGO.US) may remain the strongest existing competitor in custom chips, while standardized AI accelerators led by NVIDIA Corporation and AMD are also competing for the same spending budgets. Marvell's broad interconnect portfolio offers multiple growth paths, but winning business does not necessarily guarantee pricing power, nor does it necessarily achieve the margins required by BofA's earnings assumptions.
In addition, the Alphabet Inc. Class C agreement also deserves careful scrutiny. The much-discussed $120 billion figure involves a procurement-linked framework, not guaranteed orders. As Peace Longe pointed out when raising questions, under the warrant, nearly 59 million shares could be issued, bringing potential dilution alongside commercial opportunity.
Valuation is another pressure point. BofA's price target corresponds to about 33x calendar 2028 earnings, clearly above the 26x historical median it cites. Therefore, this high price target means investors need to continue granting Marvell a premium. A growth slowdown could hurt both earnings expectations and the valuation multiple applied to those earnings.
Therefore, when assessing the October 6 analyst day, the key focus should be on delivery timelines, margin expectations, and the ramp progress of customer projects. After such a large rally, merely expanding the market size forecast can provide only limited confidence.
At that point, the strongest confirmation signal may be clearer evidence that sales growth can sustainable actual earnings per share and cash flow per share.
Related Articles

HK Stock Market Move | BAIGE DIGITAL(02672) rose nearly 14% in morning trading. BAIGE DIGITAL appeared at the CIFIT, with MaaS business revenue surging nearly 580% in the first eight months.

HK Stock Market Move | UBTECH ROBOTICS (09880) rose over 3% during intraday trading. The Liuzhou industrial humanoid Siasun Robot&Automation super smart factory has officially commenced production, marking a new stage of mass production for the domestic industry.

Founder: Control over large model platforms and valuation restructuring, with key focus on the real TCO of domestic accelerators, etc.
HK Stock Market Move | BAIGE DIGITAL(02672) rose nearly 14% in morning trading. BAIGE DIGITAL appeared at the CIFIT, with MaaS business revenue surging nearly 580% in the first eight months.

HK Stock Market Move | UBTECH ROBOTICS (09880) rose over 3% during intraday trading. The Liuzhou industrial humanoid Siasun Robot&Automation super smart factory has officially commenced production, marking a new stage of mass production for the domestic industry.

Founder: Control over large model platforms and valuation restructuring, with key focus on the real TCO of domestic accelerators, etc.

RECOMMEND





