Bank of America Double-Upgrades Intel to Buy, Sets $135 Price Target
Bank of America analyst Vivek Arya bypassed Neutral and double-upgraded Intel to Buy on June 11, lifting his price target from $96 to $135—a call that projects more than $6 in earnings per share by 2030 on the back of a recovering CPU business and a growing foundry operation.
Intel closed Thursday at $124.57, up $7.61, after Bank of America analyst Vivek Arya did something notable even by Wall Street standards: he skipped a rating entirely. Arya moved Intel from Underperform directly to Buy on June 11—bypassing Neutral—and raised his price target from $96 to $135. The jump represents a 40.6% increase in a single recommendation, on a stock that has already gained roughly 210% in 2026.
\n\nThat 210% makes 2026 Intel's best year since 1975. The stock has been recovering on a combination of product cycle optimism, CEO Pat Gelsinger's multi-year foundry bet showing early returns, and a broader market re-rating of semiconductor companies as AI infrastructure spending stays elevated. Arya's upgrade, coming from a former skeptic, added momentum to all three narratives at once.
\n\nThe financial case rests on two revenue lines Arya believes the market currently undervalues: CPUs and foundry services.
\n\nThe Revenue Projections
\n\n| Segment | 2030 Revenue Target | Key Assumption |
|---|---|---|
| Server CPU | >$40 billion | ~25% of estimated $170B total addressable market |
| Foundry services | >$45 billion | 14A process node; potential clients including Apple and MediaTek |
| EPS by 2030 | >$6/share | Prior estimate: $3–4 |
The server CPU number assumes Intel recaptures roughly 25% of a $170 billion total addressable market by 2030—a market Arya expects to grow as agentic AI applications drive server spending beyond traditional workloads. The foundry projection depends heavily on landing and retaining anchor clients. Reports have mentioned Apple and MediaTek as potential customers, and Intel has announced a manufacturing partnership with Cadence Design Systems on the 14A node. None of those client relationships are publicly contracted at volume.
\n\nThe competitive landscape adds pressure from unexpected directions. Nvidia is actively pitching its own Vera CPU architecture for AI server deployments, complicating any single vendor's claim to the CPU opportunity. Intel's bet is that general-purpose CPUs remain the core compute layer even as accelerators multiply around them.
\n\nPut the earnings together: Arya projects Intel earning more than $6 per share by 2030, up from his prior estimate of $3–4. At roughly 100 times forward expected earnings—where the stock trades today—that trajectory either justifies the current price or the multiple has run ahead of the execution still required. Those are genuinely different outcomes.
\n\nThree Risks, and the Math Behind the Target
\n\nThe upgrade named three risks explicitly: Arm-based CPU competition, a potential slowdown in AI capital expenditure, and execution risk on the 18A and 14A manufacturing process nodes. The 18A node has been delayed before. If either process node slips again, the foundry client pipeline could thin before it was ever thick.
\n\nArya's $135 target implies about 8.4% upside from Thursday's close. That is a modest headline number for a Buy-rated stock—but the real argument is not about the next quarter. It is about whether Intel's earnings power by 2028–2030 justifies holding a stock trading at roughly 100 times near-term expected earnings today. An investor who buys at current levels is betting that the foundry clients materialize, the nodes deliver, and the CPU recovery holds. Each of those has happened at Intel before. Each has also stalled before, sometimes in the same year.
