Citi Global TMT Conf. 2025: Intel CFO on U.S. Gov’t & SoftBank Investments, Foundry Strat, Node Roadmap, and AI

Key Takeaways

TL;DR: Intel CFO Dave Zinsner highlighted major equity infusions (U.S. gov’t, SoftBank) to shore up the B/S, strict CapEx discipline (esp. for 14A node), and greater transparency via foundry/products separation. Margin recovery into the 40s (%) is a near-term target, with 18A as a long-term innovation driver. Merchant foundry wins on new nodes are critical for future CapEx. CPU/server portfolios still lag; major org restructuring underway under CEO Lip-Bu Tan.

1. Equity Investments & Fin. Position

• Gov’t Stake & Grant Swap
  • Swapped $5.7B pending US grants (prev. uncertain) + $3B+ Secure Enclave (w/ contingencies) for direct US gov’t equity stake.
  • Zinsner: “Eliminated all uncertainty around those grants... exchanged for equity stake... supports our B/S.”
  • Implication: Upfront cash, fewer restrictions/milestones, more op. flexibility, improved de-leveraging, liquidity for CapEx.
  • No governance change: “No governance... no op. input... they’ll vote w/ board recs.”
• SoftBank Investment
  • Masa Son (SoftBank) invested $2B, citing confidence in Lip-Bu Tan and AI/foundry exposure.
  • “Excited to have them as an investor.”
  • Add’l capital raise: $1B Mobileye stock sale, $3.5B from Altera close this Q.
  • Plan to retire $3.8B maturing debt in 2025 w/o refi due to these inflows.
• Potential for Ext. Investment
  • Foundry as sub enables future outside investment. “Could see minority foundry stake sale down the road,” but not near-term.

2. Mfg. Strat: CapEx, Node Roadmap, Customer Discipline

• Node Investments & Go/No-Go
  • CapEx Discipline: No 14A capacity build w/o 3rd-party customer commits (“no more blank checks”).
  • Internal demand (esp. 18A) strong, but 14A needs ext. demand for ROI.
  • 14A higher cost: High NA EUV = higher wafer cost. “Investment more significant, drives up wafer req.”
  • Timeline: “2026, we’ll know how 14A customer traction/viability looks.”
• Ext. Mfg. (TSMC)
  • Current mix: ~70/30 internal/ext. by wafer vol.; “30 comes down, but still high.”
  • Flex model: “Smart Capital” leverages ext. foundries to manage demand swings, avoid CapEx for underutilized cap.
  • Major upcoming prods (Lunar Lake, Arrow Lake) mostly/all TSMC.
  • Long-term: TSMC “great partner forever.”
• Foundry & Products Split
  • Op./Fin. split: P&Ls now separated. Product org charged for test/wafers; foundry focused on ASP/wafer cost.
  • ERP split: Dedicated ERP for foundry/products by end-2027.
  • Shift to customer-like rel. between internal design/mfg.

3. Product & Tech Roadmap

• Process Node Update
  • 18A ramp: “Panther Lake, first SKU by EOY, ramping thru 2025... Peak vol. ~2030.”
  • “18A is a workhorse node.”
  • 14A timing: For Intel/ext. customers, “2028-2029” for prod intro. Backside power key feature.
• Merchant Foundry Customers
  • Missed 1st wave for 18A ext. due to yield ramp, but DoD committed; engagement ongoing.
  • 14A PDK/customer engagement in “early stages”; earliest decisions/announcements 2026+.
• Competitive Positioning
  • CPUs: “Work to do on client side” (desktop lagging, Nova Lake/18A to improve).
  • Data center: “Need design improvements for full competitiveness.”
  • AI: CEO Lip-Bu Tan focus; strat unveil soon. Jaguar Shores upcoming AI prod.