BofA double-upgrades Intel from Underperform to Buy, raising the target price to $135
AI summary card
BofA double-upgrades Intel from Underperform to Buy, raising the target price to $135
The report believes Intel's long-term visibility in server CPUs, agentic AI CPUs, and external foundry opportunities has improved, with CY30E EPS power potentially reaching $6+, supporting a higher valuation.
- The rating was double-upgraded from Underperform to Buy, and the target price was raised from $96 to $135, versus the current price of $107.04.
- BofA expects Intel's server CPU sales to reach about $40bn+ by CY30E, representing about 25% of the $170bn CPU TAM.
- Total IDM CY30 EPS power was increased from the previous $3-4 to $6+, and the valuation was changed to 25x CY30E EPS power discounted back two years to CY28.
- External foundry opportunities include Apple M-Series wafers, MediaTek TPU wafers, Terafab IP/packaging, ARM-based server CPUs, and edge AI expansion into Client.
- Intel's ownership ratio among SPX semiconductor and AI infrastructure stocks is only 16%, relatively low versus its roughly $540bn market cap, and broader ownership could become a stock catalyst.
Report interpretation
Overview
This is a rating-change report on Intel by Bank of America. The report double-upgrades Intel from Underperform to Buy, based primarily on expanding server CPU demand, incremental CPU TAM from agentic AI, improved opportunities in external foundry wafers and packaging, and Intel's transition to a more sustainable integrated Product/Foundry IDM model.
Core views
The core view is that Intel's long-term earnings power was previously underestimated by the CY28 sum-of-parts method. BofA now values the company using CY30E EPS power of about $6.24 and a 25x P/E framework, discounted back two years to CY28, arriving at a $135 target price. The report believes server CPU sales can reach about $40bn+ by CY30E, while visibility into the external foundry ecosystem is improving, though product and foundry execution remains critical.
Analysis framework
The report develops its thesis along three lines: first, it מחדש拆分 the CY30 global data center systems TAM and CPU TAM to assess the increment from traditional CPUs, AI compute/head node CPUs, and agentic AI node CPUs; second, it switches Intel's valuation methodology from a CY28 sum-of-parts to a full-IDM long-term earnings power framework; third, it uses peer ownership comparisons to show that Intel's current institutional ownership is low, and that broader ownership could drive stock upside.
Methodology notes
Re-rating Intel based on long-term IDM earnings power
The $135 target price is based on about $6+ of CY30E EPS power at 25x, discounted back two years to CY28 to reflect the longer-term nature of the server CPU and external foundry opportunities.
Breaking down CPU demand by server architecture
The report breaks down the CPU value within the roughly $2.1Tn CY30 global data center systems TAM into about $30bn for traditional CPUs, about $70bn for AI compute/head node CPUs, and about $70bn for agentic AI node CPUs, for a total CPU TAM of about $170bn.
Low ownership could drive a broader ownership catalyst
Intel's ownership ratio among semiconductor and AI infrastructure stocks is only 16%; although up about 300bps month over month, it remains clearly low relative to its roughly $540bn market cap. The report uses AMD's ownership broadening and stock performance as a reference.
Standardized business performance and earnings quality metrics
The report discloses BofA Global Research's iQmethod framework, used to standardize business performance, earnings quality, and valuation-related metrics, though this investment conclusion is primarily driven by long-term CPU and foundry opportunities.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- Intel Corporation / INTC.OCore covered name
- Strengths
- Expanding server CPU TAM, new CPU roles driven by AI agentic nodes, growing external foundry wafer and packaging opportunities, and a relatively low institutional ownership ratio.
- Weaknesses
- Uncertainty remains around the manufacturing roadmap and foundry execution, and the details of external foundry transactions are not yet sufficient to be fully incorporated into the base model.
- Comparison
- Relative to semiconductor and AI infrastructure companies of similar market cap, Intel's ownership ratio is low; the report believes broader ownership could drive stock upside.
- Risks
- 18A and 14A yields or ramp may fall short of expectations, CPU share loss, slower AI capex, and weakness in the PC market.
- Advanced Micro Devices / AMD.USPeer comparison and ownership-broadening reference
- Strengths
- The report notes AMD's ownership increased 1400bps year over year to 39%, while its stock rose 309% year over year, serving as a case of ownership broadening driving share price performance.
- Weaknesses
- The report does not rate AMD in this note and does not elaborate on AMD's fundamental weaknesses.
- Comparison
- Intel's current 16% ownership is significantly below AMD's 39%; BofA uses the AMD case to illustrate Intel's room for broader ownership.
- Risks
- AMD remains an important competitor to Intel in server CPU and PC CPU share.
- ARM/custom CPU ecosystemSource of competitive risk
- Strengths
- ARM-based and custom CPUs have competitive expansion potential in server and AI workloads.
- Weaknesses
- The report does not separately quantify the earnings contribution of ARM/custom CPU companies.
- Comparison
- Intel's opportunity comes from expanding CPU roles, but whether its share can reach about 25% depends on the competitive outcome versus ARM/custom solutions.
- Risks
- Intensifying ARM/custom competition could pressure Intel's server CPU share and pricing.
- External Foundry opportunitiesKey opportunity set behind Intel's valuation upgrade
- Strengths
- Opportunities include Apple M-Series wafers, MediaTek TPU wafers, Terafab IP/packaging, ARM-based server CPUs, and edge AI Client expansion; the CDNS 14A partnership helps the external IP ecosystem.
- Weaknesses
- The report says it will not change the base foundry model until specific transaction details become clearer.
- Comparison
- This opportunity moves Intel from being purely a product company closer to a full Product/Foundry IDM valuation framework.
- Risks
- If substantive external foundry customers or wafer-processing orders fail to materialize, the valuation-upgrade thesis could be impaired.
Key data
- Rating changeUnderperform -> BuyThe report describes this as a double-upgrade.
- Target price$135Raised from the previous $96.
- Current price$107.04Price disclosed on the report cover page.
- Implied upsideapproximately 26.1%Calculated based on the $135 target price and the current price of $107.04.
- CY30E EPS power$6+, approximately $6.24This is the core assumption of the new valuation framework, versus about $3-4 previously.
- CY28E EPS powerapproximately $3.47Previously about $2.25, based on the valuation methodology comparison.
- CY30E server CPU salesapproximately $40bn+About 25% of the $170bn CPU TAM.
- CY30 data center systems TAMapproximately $2.1TnOf which CPU value is about $170bn.
- AI CPU TAMapproximately $140bnAbout $70bn for AI compute/head node and about $70bn for agentic AI node.
- Non-AI CPU TAMapproximately $30bnTraditional, on-prem, and multi-tenant cloud CPU demand.
- Intel ownership ratio16%As of May 2026, up about 300bps month over month.
- Intel market capapproximately $540bnApproximately ranked No. 5 within the relevant semiconductor and AI infrastructure group.
- CY27-28E estimate revisionssales up 1-3%, pf-EPS up 9-12%From Exhibit 3 of the report.
- Valuation multiples25x CY30E EPS power; 9.1x 2028E EV/SEV/S is significantly above the historical 1.7x-4x range, which the report believes is supported by the server CPU and foundry opportunities.
Impact & implications
If Intel can deliver on higher server CPU share, 18A/14A node ramp-up, and external foundry customer wins, the market may shift from focusing on short- to medium-term manufacturing uncertainty to re-rating long-term IDM earnings power. For investors, the report positions Intel as a high-conviction, multi-year upgrade-cycle name, but returns remain highly dependent on product and foundry execution.
Risks
- Intensifying ARM/custom competition could cause Intel to continue losing CPU share.
- A slowdown in AI capital spending could affect CPU spending and server demand.
- 18A and 14A node yields, ramp-up, and manufacturing roadmap execution may fall short of expectations.
- The external foundry business may lack substantive wafer-processing customers or see key deals materialize more slowly than expected.
- The PC market is mature and may be weaker than expected, affecting Intel's largest revenue source.
- Power and cost constraints may affect data center expansion and CPU demand.
- The target price implies 9.1x 2028E EV/S, above the historical 1.7x-4x range, making it highly dependent on successful execution.
What to watch
- 18A mass-production progress, 14A node yields, and follow-up to the CDNS 14A ecosystem partnership.
- Whether external foundry opportunities such as Apple M-Series wafers, MediaTek TPU wafers, and Terafab IP/packaging convert into substantive orders.
- Whether Intel server CPU sales can advance toward about $40bn+ and roughly 25% CPU TAM share by CY30E.
- Whether AI compute/head node and agentic AI node CPU demand can support about $140bn of AI CPU TAM.
- Whether Intel's institutional ownership ratio can broaden further from 16%, bringing AMD-like inflows.
- The impact of AI capital spending, Windows 10 refresh, and AI PC demand on CPU and Client businesses.