Topic: MCU vs analog revenue mix and strategic shift to data center/AI
Key points:
June quarter: analog growth stronger than MCU; September quarter reversed (MCU better, analog worse).
Steve Sanghi confirmed a strategic shift: hired Brian McCarson (Jan 2025) to lead data center products to state-of-the-art; first device is 3-nanometer Gen 6 Switchtec.
Formed an AI business unit and increased FPGA focus; upcoming FPGA announcements and a NASA contract for next-gen space computer.
Mgmt stance: Bullish — shift aims to pull overall CAGR higher via advanced-node, high-performance products.
Q7 — William Stein
Topic: Underutilization charges, inventory write-downs, and gross margin normalization timeline
Key points:
Steve Sanghi: not comfortable forecasting a timeline; expects “substantial improvement” in next fiscal year (starting June 2026), but no absolute timeframe.
J. Bjornholt: inventory write-offs normalize faster than underutilization; both moving in right direction.
Product gross margin target is 67%+ , but long-term model is 65% ; current non-GAAP gross margin guided to 58.2% midpoint (vs 56.7% prior).
Normal utilization varies by factory (e.g., 90% in one, 75% in another); 68%+ margins in up cycle were aided by expedite charges, not standard.
Mgmt stance: Neutral — cautious on timing; underutilization will persist longer, and 65% is the current long-term target.
Q8 — Harlan Sur
Topic: Fab 2 closure cost savings and data center segment mix
Key points:
Fab 2 closure targets $90 million annual cash savings; agreement subject to closing conditions (possible Dec 2025 cash inflow, amount undisclosed).
Fab 2 costs not impacting non-GAAP gross margins currently; main drags are underutilization and inventory write-offs.
Data center & compute segment was 19% of fiscal 2025 total revenue; majority is data center (not client/PC), but exact mix not disclosed.
Mgmt stance: Neutral — progress on Fab 2 savings, but no quantification of remaining impact; segment mix not broken out further.
Steve Sanghi: three tailwinds for March 2026 and beyond — (1) distributor sell-in/sell-through gap (~$50M) to close as inventories normalize; (2) direct customers and contract manufacturers buying less than consumption; (3) June/September are seasonally strong quarters.
3-nanometer Switchtec family targets hyperscaler, enterprise OEM/ODM; total addressable market > $2 billion/year today, with >10% CAGR through 2035.
Production release expected by June 2026; first revenue in late 2026 and 2027.
Mgmt stance: Bullish — multiple structural drivers for revenue recovery; new product pipeline supports long-term growth.
Q10 — Christopher Danely
Topic: September quarter booking trends, December quarter softness, and substrate constraints
Key points:
July: one of best booking months in 3 years; August slower but better than expected; September: best booking month in 3+ years, book-to-bill positive, bookings up 10% sequentially.
December quarter softness due to customers scheduling strong bookings for January (calendar year-end balance sheet management), not December; March backlog stronger than December on key product lines.
Substrate capacity remains constrained, competing with cell phone builds (now easing); some customers delayed shipments by one quarter, impact “meaningful” but not hundreds of millions.
Mgmt stance: Neutral — demand strong but timing shift causes near-term disappointment; constraints manageable but not fully resolved.