Bank of America Sees AI Driving a New $1 Trillion Wave in Semiconductors Within Five Years

Keywords: artificial intelligence, semiconductors, Bank of America, AI data centers, memory chips, wafer fab equipment, Micron, Intel, Marvell, Arm

Introduction

As concerns grow that the “AI bull market” may have peaked, Bank of America has issued a sharply more optimistic long-term outlook for the semiconductor industry. According to the firm’s latest research, artificial intelligence could help the chip sector generate an additional $1 trillion in sales within just five years—an achievement that took the industry roughly half a century to reach the first time.

Led by analyst Vivek Arya, the BofA team argues that the next phase of AI adoption is not merely a software story, but a broad-based infrastructure cycle that will reshape demand across the entire semiconductor value chain. Their thesis rests on five key growth engines that could accelerate industry revenue far beyond current expectations.

Five Structural Drivers Behind the Next Semiconductor Boom

1. AI Data Centers: A Multi-Year Capacity Expansion

The most powerful driver, according to BofA, is the explosive growth of AI data center systems. The firm estimates that the total addressable market for this segment could rise from $273 billion in 2025 to approximately $1.7 trillion by 2030. That scale of expansion reflects not only the continued deployment of large language models, but also the rise of inference-heavy workloads, agentic AI applications, and enterprise-level AI integration.

Unlike earlier technology cycles, this is not a one-time spending burst. It is an infrastructure buildout that requires repeated waves of investment in compute, networking, memory, and power management. In BofA’s view, AI data centers are becoming the central physical layer of the digital economy.

2. Memory Chips Supported by Long-Term Supply Agreements

The second major catalyst is memory demand, particularly in high-bandwidth memory and other advanced products used in AI accelerators. BofA believes the memory market is entering a period of stronger pricing power and improved resilience, supported by long-term supply agreements.

This matters because memory has historically been one of the most cyclical parts of semiconductors. However, the AI era is changing that dynamic. With hyperscalers and chip designers seeking guaranteed supply, the sector is moving toward a more disciplined and structurally tighter market. BofA even expects memory chip sales to surge sharply in 2026, potentially posting growth approaching 300% year over year.

3. Semiconductor Capital Equipment and Reshoring

A third driver is the rising complexity of chip manufacturing, which in turn supports demand for semiconductor capital equipment. As nodes become more advanced and packaging technologies become more sophisticated, foundries and memory makers must invest heavily in wafer fab equipment.

At the same time, reshoring efforts in the United States and other regions are reinforcing that trend. Governments and companies alike are pushing for more diversified and secure supply chains, which means additional fab construction, tooling, and process investment. BofA has therefore raised its expectations for wafer fab equipment spending substantially.

4. Analog Chips Benefit From Surging AI Compute

The fourth growth engine is the analog chip market. While AI headlines often focus on GPUs and CPUs, the infrastructure behind AI depends on a wide range of analog components, including power management, signal conversion, and connectivity solutions.

As computing density rises, power efficiency and system stability become more important. This increases the strategic value of analog semiconductors, especially in data centers, networking gear, and edge devices. BofA sees this as an underappreciated area of upside in the AI cycle.

5. Agentic CPUs and the Next Server Architecture Shift

The fifth factor is the emerging demand for agentic central processing units. BofA estimates that this space could create a $170 billion server market opportunity across both x86 and ARM architectures.

This is a notable point because it suggests that AI growth will not be confined to specialized accelerators. Instead, it may trigger a broader redesign of server architecture, where CPUs, accelerators, memory, and networking all evolve together to support autonomous AI agents and more complex workloads.

A Broader Upgrade to the Semiconductor Outlook

In line with this more bullish framework, Bank of America also lifted its forecast for the global semiconductor industry’s total market size in 2030 from $2.3 trillion to $2.7 trillion. That implies a compound annual growth rate of 28% from 2025 to 2030, an exceptionally strong trajectory for such a large industry.

The revision underscores a key point: AI is no longer just a theme for a few select companies. It is becoming a system-wide demand driver that could extend the semiconductor supercycle far longer than many investors initially expected.

BofA also raised its outlook for wafer fab equipment spending. It now expects WFE expenditure to reach $250 billion in 2028, up from a prior estimate of $203 billion, and to climb further to $292 billion by 2030. This reflects both the intensity of AI-related capacity expansion and the growing technological complexity of advanced chip production.

Stock Calls Reflect the New AI Reality

The report was not limited to industry forecasts. BofA also adjusted target prices for several semiconductor stocks, signaling where it sees the strongest beneficiaries of the AI investment wave.

Micron: The Most Aggressive Upgrade

The most notable move was on Micron Technology, whose target price was raised from $950 to $1,500. BofA cited strong demand for high-bandwidth memory and a structural supply constraint that could persist through 2028.

This view aligns with broader market sentiment. Deutsche Bank previously raised its Micron target from $1,000 to $1,500, while Citi increased its target from $840 to $1,200. The consensus shift suggests that investors increasingly view memory as a strategic bottleneck in the AI ecosystem rather than a commoditized segment.

Intel: Confidence in Servers and Foundry Ambitions

Intel’s target price was lifted from $135 to $160, marking the second major upgrade in just two weeks. On June 11, BofA had already upgraded Intel from “underperform” to “buy” and raised its target from $96 to $135.

The latest increase reflects growing confidence in Intel’s server CPU business and its foundry strategy. If AI demand broadens beyond GPUs and into traditional compute infrastructure, Intel could regain relevance in key parts of the enterprise and cloud stack.

Marvell and Arm: Connectivity and Architecture Tailwinds

Marvell Technology saw its target raised from $240 to $365, driven by strong AI connectivity demand. This makes sense in a market where bandwidth, interconnect, and data movement are increasingly critical performance constraints.

Arm Holdings also received a higher target, rising from $335 to $460, although BofA kept a neutral rating. The firm appears to believe that Arm’s architecture has meaningful upside, but that much of the opportunity is already priced in.

Equipment Leaders Also Benefit

BofA raised targets for several semiconductor equipment names as well. Applied Materials moved from $540 to $720, Lam Research from $330 to $480, and KLA from $210 to $317. These revisions reflect the bank’s view that the AI cycle is translating directly into higher capital intensity across manufacturing and metrology.

A Rare Exception: Axcelis Technologies

Among the broadly positive adjustments, Axcelis Technologies stood out as the exception. BofA lifted its target from $130 to $156 but maintained an underperform rating, arguing that the stock already reflects much of the upside from its pending merger with Veeco.

This caution is important. Even in a strong AI-led capital cycle, not every semiconductor name offers equal risk-reward. Valuation discipline remains essential, especially after a strong rally across the sector.

Conclusion

Bank of America’s latest semiconductor outlook sends a clear message: the AI revolution is still in its early stages, and its economic footprint may be far larger than many investors assume. Rather than signaling the end of the cycle, current market volatility may simply reflect the transition from an initial hype phase to a deeper, infrastructure-driven expansion.

With AI data centers, memory, capital equipment, analog chips, and next-generation server CPUs all contributing to growth, the semiconductor industry could be on track for one of the most significant revenue expansions in its history. For investors, the implication is straightforward: the AI story is no longer just about model performance or software adoption. It is about the entire hardware stack—and that stack may still have years of runway ahead.