AI power expansion drives target price increases for European semiconductors, with IFX seen as the best-positioned beneficiary

Overview

This report focuses on the opportunities for European semiconductor companies in the AI power cycle, especially comparing the positioning of Infineon Technologies AG and STMicroelectronics NV in data-center power, SiC, GaN, CPU/GPU power management, and power infrastructure. The report’s core conclusion is that AI benefits are not limited to GPUs; power analog, SiC/GaN, VRMs, and infrastructure power conversion will also expand significantly, benefiting European power semiconductor companies.

Core views

BofA believes IFX is best positioned to capture the expansion of AI power infrastructure, benefiting from its depth in power semiconductors, GaN portfolio, and exposure to PSU, PCS, SST, SSCB, and ESS. STM, while having a strong automotive foundation in the broader SiC market and more leverage to areas such as optics and LEO, still has relatively low penetration in data-center SiC. The report raises target prices and some earnings forecasts for both companies, but IFX earns a relative premium due to its leadership in AI power.

Analysis framework

The report converts AI compute demand into power deployment, analog semiconductor TAM, device TAM, and company revenue opportunities, then combines SiC/GaN penetration, CPU inference power demand, market-share assumptions, segment revenue upgrades, and EV/EBITDA valuation multiples to derive target-price revisions for IFX and STM.

Methodology notes

• Market sizingTAM expansion framework

  Derives analog semiconductor market size from AI deployment power, data-center, and power-infrastructure demand.

  The report estimates about 233GW of deployment by 2030E and expands the AI analog market from $7.9bn to $27bn to assess the revenue opportunity for European power semiconductor companies.

• Competitive positioningShare and product portfolio comparison

  Compares IFX and STM’s relative exposure to niche opportunities such as SiC, GaN, VRMs, optics, and LEO.

  IFX is more concentrated in AI power and data-center power devices, while STM relies more on its automotive SiC base and adjacent growth areas such as optics and LEO.

• Valuation methodsEV/EBITDA multiple method

  Adjusts target prices using FY28E EV/EBITDA multiples combined with revenue and gross-margin forecasts.

  IFX’s target price is based on 15.0x FY28E EV/EBITDA, while STM’s target price is based on 11x FY28E EV/EBITDA, both reflecting a higher AI revenue mix while still remaining below some peers.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Infineon Technologies AG (IFNNF / IFNNY)

  Preferred beneficiary of AI power and European power semiconductors.

  Strengths

  Strong depth in power semiconductors, expanding GaN portfolio, and exposure to AI power infrastructure such as PSU, PCS, SST, SSCB, and ESS; data-center SiC share and VRM share are expected to rise.

  Weaknesses

  Relatively less well positioned than some competitors in rapid automotive EV SiC adoption scenarios; high CAPEX may weigh on cash conversion.

  Comparison

  Versus STM, IFX is more concentrated and more of a leader in AI power infrastructure and data-center power devices, and therefore receives a higher valuation multiple and a Buy/top pick rating.

  Risks

  EV and ADAS adoption slower than expected, faster-than-expected SiC adoption in EVs with IFX at a relative disadvantage, order delays following double ordering caused by chip shortages, high CAPEX, and euro appreciation against the U.S. dollar.

• STMicroelectronics NV (STMEF / STM)

  A beneficiary of the AI power cycle, but rated Neutral.

  Strengths

  Has high share in the broader SiC market, a strong automotive foundation, and greater leverage to the fast-growing optical market, Silicon Photonics, and LEO satellite opportunities.

  Weaknesses

  Still underpenetrated in data-center SiC, currently has low VRM share, and is less concentrated than IFX in AI power positioning.

  Comparison

  Compared with IFX, STM receives a target-price increase but a lower valuation multiple; the report believes its ability to gain AI power share is weaker than IFX’s.

  Risks

  Macro weakness and destocking pressure revenue and earnings, loss of Apple design wins or pricing pressure, EUR/USD rising above 1.20, potentially value-destructive M&A, and memory shortages affecting consumer and automotive demand.

Key data

• AI-related power deploymentabout 233GW by 2030EUsed to support the assumption of AI power semiconductor market expansion.
• AI analog market sizefrom $7.9bn to $27bn, 28% CAGRThe report estimates the expansion path from the current market to 2030E.
• Forecast data-center SiC shareIFX about 37%, STM about 15% by 2027/28EThe report believes IFX will lead STM in data-center SiC.
• SiC growth rateabout 73% CAGR 2025-30EViewed as one of the fastest-growing subsegments.
• CPU semiconductor TAMfrom $671mn in 2025 to $3.1bn by 2030E, 36% CAGRGrowth driven by inference demand and higher power-management content per machine.
• IFX data-processing voltage regulator shareabout 2% in 2024 to about 7% in 2025The report says IFX has increased share in relevant power-management links.
• IFX server CPU power revenue forecast€410-760mn 2026-28EDriven by increasing VRM complexity and migration to vertical power delivery.
• IFX target price revisionfrom €70/US$82 to €95/US$110Mainly driven by AI server revenue, PSS mix, and GIP upgrades.
• STM target price revisionfrom €49/US$57 to €63/US$73Reflects upgrades to AI/data-center revenue, gross margin, and P&D forecasts.
• Valuation multiplesIFX 15x FY28E EV/EBITDA; STM 11x FY28E EV/EBITDAAbove their respective historical ranges but below some AI-exposed peers.

Impact & implications

If AI data-center power architectures evolve toward higher power density, higher voltage, and more complex power management, the value of power semiconductors may spill over beyond GPU/CPU platforms into analog, SiC, GaN, VRM, and power infrastructure. The report implies the market may be underestimating IFX’s ability to gain share from scaled AI power deployment, while STM’s opportunity depends more on expansion into optics, LEO, and non-automotive scenarios.

Risks

• AI power deployment, data-center capital expenditure, or AI server demand may come in below expectations.
• The pace of SiC/GaN adoption in data centers and power infrastructure may fall short of expectations.
• Slower demand for EV, ADAS, and automotive semiconductors could affect the non-AI core business of IFX and STM.
• A stronger euro against the U.S. dollar would pressure earnings exposed to U.S.-dollar-linked revenue or cost structures.
• Chip shortages, double ordering, destocking, and weakening macro demand could lead to delayed orders or revenue downgrades.
• Company-specific risks include high CAPEX, loss of design wins, pricing pressure, value-destructive M&A, and memory shortages.

What to watch

• Whether the assumption of about 233GW of AI power deployment by 2030E is realized.
• Whether the AI analog market can expand at a 28% CAGR to $27bn.
• Whether IFX’s share in data-center SiC and VRM continues to rise.
• Whether CPU inference demand, vertical power delivery, and greater power-management complexity add incremental semiconductor content.
• STM’s progress in Silicon Photonics, LEO, and AI power P&D revenue.
• Whether upgrades to IFX’s PSS, GIP, and AI server revenue can translate into gross margin and EPS delivery.
• The EUR/USD trend and its impact on the revenue and earnings of both companies.