BofA Upgrades Intel to Buy, Projects $170B Server CPU Market by 2030

For investors, the semiconductor story has been overwhelmingly GPU-centric—until now. Bank of America’s upgrade of Intel (INTC) and a bold $170 billion CPU market forecast signal that the next leg of the AI trade could belong to a chip giant that many had left for dead. With the emergence of agentic AI, the financial community is being forced to reassess which processors will power the data centers of the future, and that rebalancing could produce significant alpha for early movers.

In a notable shift for the semiconductor landscape, Bank of America has delivered a resounding signal that the AI narrative is about to broaden beyond graphics processors. Analyst Vivek Arya upgraded Intel from underperform to buy and lifted the price target to $135 from $96, while simultaneously raising the bank's forecast for the server CPU market to nearly $170 billion by 2030, up from a prior estimate of $125 billion. The catalyst is the expected rise of agentic AI—systems that move beyond simple chatbot responses to planning, reasoning, retrieving information, and executing multi-step tasks. This evolution, Arya argues, will reposition central processing units as a primary computing engine, rather than the supporting role they played during the training-heavy first phase of AI.

“

Bank of America’s upgrade of Intel (INTC) and a bold $170 billion CPU market forecast signal that the next leg of the AI trade could belong to a chip giant that many had left for dead.

The upgrade marks a rare piece of positive news for Intel, a company that has been battered by manufacturing delays, market-share erosion, and deepening investor skepticism about its relevance in an AI world dominated by Nvidia's GPUs. For years, the AI hardware story was almost singularly focused on parallel processing accelerators, propelling Nvidia, AMD, Broadcom, and Marvell into the spotlight. CPUs, while essential for data preparation and orchestration, were largely invisible to the investment narrative. That is now poised to change.

Agentic AI represents a paradigm in which models must continuously engage in logical reasoning, access external tools, and chain together complex workflows. These tasks are less about raw matrix multiplication and more about the control, branching, and decision-making at which modern server CPUs excel. The shift does not eliminate the need for GPUs, but it dramatically expands the addressable market for high-performance x86 and alternative processors. Bank of America's revised $170 billion forecast reflects this expectation, implying a compound annual growth rate well above that of the broader semiconductor industry.

For Intel, the upgrade is strategically significant beyond the immediate stock price boost. The company's Xeon server processors, which have faced fierce competition from AMD's EPYC line, stand to benefit from a rising tide. Even if Intel's market share has eroded, the sheer expansion of the total addressable market gives it a path to revenue recovery. Moreover, the upgrade directly challenges the bearish thesis that Intel's legacy CPU business is a secular decline story. If agentic AI workloads become pervasive, data center operators may need to rebalance their infrastructure spending, allocating more to CPU-heavy nodes alongside GPU clusters.