Sign in

    Intel Corp (INTC)

    Q1 2024 Summary

    Published Jan 10, 2025, 5:10 PM UTC
    Initial Price$50.81December 28, 2023
    Final Price$44.17March 28, 2024
    Price Change$-6.64
    % Change-13.07%

    What went wrong

    • Weaker than expected demand across cloud, enterprise, and geographies is leading to low revenue growth in the near term.
    • Intel Foundry Services is operating at a loss, with operating losses expected to increase through 2024, and profitability not anticipated until midway between now and 2030.
    • Delays in launching new server products like Granite Rapids, with production release in Q3 2024, and time needed for customer adoption could delay market share gains.

    Q&A Summary

    1. Gross Margin Outlook
      Q: Can you discuss the gross margin drop in Q2 and outlook?
      A: Gross margins had an upside in Q1 due to better sell-through on previously reserved material, pulling some benefit from Q2 into Q1. The Q2 drop is due to increased startup costs. However, they expect good fall-through in Q3 and Q4, lifting gross margins from current levels. They anticipate better gross margins in 2025 than in 2024, aiming for mid-50s gross margins by mid-decade and ultimately reaching 60%. Cost improvements from better decision-making and changes in reserving practices will reduce volatility in gross margins.

    2. Demand Outlook and H2 Growth
      Q: What's behind weaker near-term demand and H2 growth confidence?
      A: The market was weaker across cloud, enterprise, and geographies. Despite this, they see strength in H2 driven by unique product positions, AI PC demand, a Windows upgrade cycle, and strong interest in Core Ultra processors. They're experiencing a strong AI PC outlook and improved competitiveness in data center products. All businesses are expected to grow in the second half, with momentum going into 2025.

    3. Server Roadmap and Market Share
      Q: How will new server products impact market share?
      A: They are stabilizing and improving competitiveness in the data center with new products like Sierra Forest and Granite Rapids on Intel 3 process. ASPs are increasing due to higher core counts. They expect to regain share as they end the year and move into 2025, with strong products and aggressive ramps with customers.

    4. AI Accelerator Revenue and Roadmap
      Q: Update on Gaudi and Falcon Shores AI accelerators revenue?
      A: They expect over $500 million in Gaudi revenue this year, mostly in the second half, indicating strong momentum into 2025. Over 20 customers are embracing Gaudi 2 and 3. Falcon Shores will build on this momentum, arriving late next year, combining Gaudi 3 performance with full programmability.

    5. AI PCs and Supply Constraints
      Q: How are supply constraints affecting AI PC outlook?
      A: There is strong demand for AI PCs (Core Ultra processors), with customers requesting additional supply beyond commitments. Supply constraints are in the back-end wafer-level assembly capacity, which they are working to expand. AI PCs are creating new use cases and an upgrade cycle, leading to higher ASPs and potential TAM expansion. They expect every PC to become an AI PC over time.

    6. Foundry Business Operating Loss and Revenue
      Q: How should we model foundry operating losses and revenue?
      A: Foundry operating losses will increase this year due to startup costs, reaching around $2.4 billion in Q2 and remaining at that level throughout the year. Improvement is expected in 2025, aiming to reach breakeven midway between now and 2030. External wafer volumes on Intel 4 and 3 will become more prominent next year.

    7. Role of CPU in Gen AI Servers
      Q: Does CPU matter in gen AI servers with accelerators?
      A: Yes, CPUs play an important role, especially in enterprise use cases where data resides in x86 databases. Intel sees opportunities as CPUs handle real-time data, security, networking, and management, integrating with accelerators for AI workloads. They are focusing on unlocking enterprise AI with Xeon CPUs plus Gaudi accelerators, enabling new applications like 70-billion-parameter models running natively on Xeon.

    8. Insourcing of Tiles and Process Leadership
      Q: Any flexibility to bring back external tiles earlier?
      A: There is limited flexibility since process node decisions are made early in product planning. The peak of external tiles is this year and next, with a significant shift back to Intel factories starting late next year. They expect to insource a couple of fabs' worth of capacity by 2026 and beyond, improving wafer ASPs, volume, and margins.

    9. Altera and NEX Business Growth
      Q: What's driving the expected ramp in Altera and NEX?
      A: Altera is expected to exit the year at a $2 billion run rate, recovering from inventory digestion and normalizing to end consumption. NEX (Network and Edge Group) will grow as it overcomes inventory adjustments, with strength in AI networking products, IPUs, and edge AI as they enter the second half and into 2025.

    Related Publications