Intel (INTC) Q2 2025: Foundry Losses Widen to $3.2B as Utilization Falls Below Break-Even

Intel’s foundry business posted a $3.2 billion operating loss, spotlighting the cost of excess capacity and underutilized assets even as management pivots to a more disciplined, customer-driven approach. CEO Lipu Tan is accelerating organizational streamlining and CapEx restraint, but the path to foundry profitability and competitive server share remains long. Investors should watch for execution on 18A ramp, AI platform strategy, and further CapEx discipline into 2026.

Summary

  • Foundry Losses Deepen: Excess capacity and impairment charges underscore the challenge of turning foundry into a profitable business.
  • Organizational Overhaul Accelerates: Management is slashing headcount and management layers to restore agility and reduce cost.
  • CapEx Discipline in Focus: Future investment will be strictly tied to customer commitments and internal product milestones.

Performance Analysis

Intel’s Q2 results reveal a company in the midst of a foundational reset as it contends with legacy overbuild and shifting demand patterns. The headline foundry loss of $3.2 billion, worsened by $800 million in impairment charges, highlights the drag of underutilized assets and the cost of past capacity investments made ahead of demand. Core Intel Products revenue performed above expectations, but gross margin was compressed by both restructuring costs and the early ramp of new nodes.

Client Computing Group (CCG) saw continued refresh demand, supported by Windows 10 end-of-life and AI PC mix shift, while Data Center & AI (DCAI) revenue declined sequentially but outperformed internal forecasts due to strong host CPU demand and the ramp of Granite Rapids. CapEx discipline is now central, with 2025 gross CapEx guided at $18 billion, but management signaled a downward trajectory for 2026 as digestion of prior investments continues. Operating cash flow remains negative, and the company is monetizing non-core assets to bolster liquidity.

  • Foundry Operating Losses Expand: Driven by underutilization, impairments, and lack of external volume, foundry remains a material drag.
  • Gross Margin Under Pressure: Non-recurring charges, Lunar Lake memory mix, and early Panther Lake ramp diluted profitability.
  • Server Share Still at Risk: Despite improved product cadence, competitive gaps and ASP (average selling price) declines persist in key segments.

The quarter’s results underscore that Intel’s transformation is far from over, with execution on cost, capacity, and product delivery now the fulcrum for future upside.

Executive Commentary

"Our factory footprint has become needlessly fragmented. Going forward, we will grow our capacity based solely on the volume commitments and deploy CapEx lockstep with the tangible milestones and not before."

Lipu Tan, Chief Executive Officer

"We recognized approximately $800 million of non-cash impairment and accelerated depreciation charges related to excess prior generation tools for which we couldn't find reuse and approximately $200 million of one-time period costs. These charges resulted in Q2 gross margin of 29.7% and EPS of minus 10 cents."

David Zinsner, Chief Financial Officer

Strategic Positioning

1. Foundry Reset and CapEx Discipline

Intel is pivoting its foundry strategy to a customer-commitment model, abandoning the “build it and they will come” approach that led to excess capacity. Strategic projects in Germany and Poland have been halted, and Ohio construction slowed, with future CapEx tied to tangible customer volume. This marks a sharp break from prior years’ speculative capacity expansion.

2. Organizational Streamlining and Accountability

Management is aggressively reducing complexity, with headcount on track to reach 75,000 by year-end and management layers cut by 50 percent. The return-to-office mandate and focus on retaining top talent are intended to restore agility and execution speed, with direct CEO oversight on all major chip tape-outs to enforce discipline.

3. Product Roadmap and AI Strategy

Penta Lake and Panther Lake launches anchor the x86 franchise, with 18A process ramping as the foundation for three future generations. In AI, Intel is shifting from a silicon-centric to a full-stack approach, aiming to differentiate in inference and agentic AI workloads by combining x86 CPUs, xe GPUs, and new system software.

4. Balance Sheet and Asset Monetization

Intel is monetizing non-core assets, including a $900 million Mobileye stake sale and the pending Altera transaction, to shore up liquidity and begin delevering. The company is targeting operating expense reductions to $17 billion in 2025 and $16 billion in 2026, with further CapEx cuts planned as prior investments are digested.

Key Considerations

Intel’s Q2 marks a turning point in operational and capital discipline, but execution risk remains high as the company navigates a multi-year turnaround.

Key Considerations:

  • Foundry Losses Remain a Drag: Persistent operating losses and lack of external customer traction highlight the uphill battle in foundry turnaround.
  • CapEx Overhang Limits Flexibility: Large construction-in-progress and prior overbuild restrict near-term ability to reallocate capital efficiently.
  • Server Share and ASP Pressure: Competitive gaps in high-end server and multi-threading capabilities continue to weigh on market share and pricing.
  • AI Platform Ambitions Unproven: Full-stack AI strategy is in early stages, with execution on software and systems integration still to be demonstrated.

Risks

Intel faces significant execution risk in right-sizing capacity and restoring foundry credibility, with the risk of further impairment charges if external demand fails to materialize. Competitive threats from ARM and continued ASP erosion in server CPUs could prolong share losses. Macro uncertainties, such as tariffs and regulatory shifts, add further unpredictability to both demand and cost structure.

Forward Outlook

For Q3 2025, Intel guided to:

  • Revenue range of $12.6 to $13.6 billion, flat to up 6% sequentially
  • Gross margin of approximately 36% (non-GAAP basis)
  • Break-even non-GAAP EPS

For full-year 2025, management maintained:

  • OpEx target of $17 billion, with 2026 target of $16 billion
  • CapEx to decline in 2026 as prior investments are digested

Management cited ongoing tariff uncertainty, capacity constraints in Intel 7, and the early ramp of new products as key variables influencing the second half. Guidance assumes a below-seasonal revenue pattern as prior quarters may have benefited from customer hedging.

  • CapEx flexibility will depend on customer volume commitments
  • Gross margin improvement contingent on mix and ramp efficiency

Takeaways

Intel’s turnaround hinges on capital discipline, execution on new nodes, and restoring foundry credibility.

  • Foundry Losses Highlight Structural Challenge: Persistent losses and asset impairments stress the need for a more demand-aligned approach and external customer validation.
  • Organizational Streamlining Key to Margin Recovery: Deep cuts in headcount and management layers are necessary but will take time to translate into sustained operating leverage.
  • Watch for 18A Ramp and AI Platform Progress: Execution on new product launches and full-stack AI strategy will determine whether Intel can regain competitive ground and attract foundry customers.

Conclusion

Intel’s Q2 underscores the cost of past overexpansion and the scale of the turnaround challenge ahead. The company is now focused on capital restraint and operational discipline, but sustained improvement will depend on execution across foundry, server, and AI initiatives.

Industry Read-Through

Intel’s retrenchment on speculative foundry capacity sends a cautionary signal to the broader semiconductor industry about the risks of building ahead of demand. The shift to customer-committed CapEx and organizational streamlining could become a template for other chipmakers facing cyclical overbuild. Persistent server CPU pricing pressure and ARM’s encroachment foreshadow continued margin compression across legacy x86 players. For the AI hardware ecosystem, Intel’s pivot to a full-stack approach highlights the growing importance of software and systems integration, but also the high execution bar for late entrants seeking to challenge entrenched incumbents.