Stacy Rasgon

Zinsner explained the write-down as stranded, incomplete match-set inventory and noted Panther Lake/Granite mix offsets the reserve tailwind, keeping margins flat.

Zinsner attributed the client beat largely to ASPs and mix, and guided client revenue roughly flat next quarter with a softer underlying market.

Dave indicated Intel client revenue will outperform the TAM decline due to pricing and customer inventory replenishment, with Q2 revenue roughly setting the run rate for H2.

Dave explained Panther Lake volume increasing 6-7x in Q2 vs Q1 but still below corporate average margins, offsetting pricing and mix benefits. Materials costs are a H2 headwind.

Dave acknowledged the miss: 6 months ago, core count was expected to increase but not units. Hyperscaler demand rapidly shifted to unit growth in H2 2025, a dynamic that could persist for several years. Intel wasn't planning supply for that level of unit demand.

Dave confirmed both segments down in Q1 due to supply constraints. Buffer inventory depleted to 40% of peak, now operating hand-to-mouth from fab output. Q1 is the trough with improvement starting in Q2.

Dave pushed back, AI PC was up double digits sequentially with ~100M unit target on track. Older-node strength was unexpected, driven by Windows refresh rather than AIPC rejection. Working with ISVs to drive AI PC application ecosystem.

Dave clarified: not adding incremental supply beyond existing CapEx plan, but volume will ramp through the year. Yields on track for year-end goal but need improvement to be fully margin-accretive, that takes all of 2026.

Dave said 2026 CapEx will be between $9B and $18B, locking in around January 2026. Assets under construction won't fully cover needs even at maintenance levels.

Lip Bu confirmed 14A timing at 2028/2029, aligned with TSMC A14. Won't deploy CapEx until customer volume commitments exist. Dave added 18A won't hit peak volumes until early next decade with opportunity for external customers in later waves.

Michelle said Panther Lake has better price-performance than Lunar Lake, strong commercial demand for AIPC remains, and the traditional ramp goes commercial-first then consumer.

Michelle attributed it to customers needing lower system price points amid macro concerns and tariff hedging. Raptor Lake has ridden cost curves down while Lunar Lake carries higher system ASP.

Dave attributed equal declines to macro uncertainty and seasonality but didn't fully explain why DCAI and NEX are as weak as client despite different dynamics.

Michelle acknowledged increased competition from new CCG entrants and committed to fighting for every socket in both client and data center markets.

Quantified both sides: a 48% to 50% baseline range restored by pricing, less a 1.5% to 2% weighted-average drag from custom silicon.

Rejected the framing and confirmed supply constraints were only part of it, but did not explain what else drove the outcome or who initiated it.

Akash acknowledged China recovery offset by Apple step-down but declined to guide September quarter handset direction specifically.

Cristiano acknowledged memory uncertainty but noted new capacity from emerging memory players and evolving dynamics between data center and edge demand.

Akash noted Q2 QTL slightly below last year. Full year has negative bias on units given supply constraints. Strong Q1 units exceeded expectations.

Akash suggested March-like seasonality for June. Not guiding beyond Q2 given uncertainty. Supply alignment will determine financial forecast for rest of fiscal year.

Akash denied pull-forward, attributing strength to normal launch timing. All major China customers launched with positive initial reception; global customers launching later in quarter.

Akash cited two drivers: mix shift toward premium tier (across all regions, not just developed) and content growth within premium tier. Both multiyear trends.

Akash confirmed normal December seasonality across all businesses, adjusted only for lower Apple share. Declined to give specific December quarter guidance but pointed to historical trend.

Akash acknowledged minor direct impact reflected in numbers but said indirect effects are difficult to predict. Not significant enough to change guidance direction.

Akash attributed small margin dip to lower-tier mix within handsets, said broader trend is healthy with no new structural information.

Cristiano firmly denied any tariff-related pull-forward, attributing demand strength to Samsung GS25 global share, China premium tier expansion, and Snapdragon 8 Elite sell-through.

Alex Rogers confirmed flat QTL guidance excludes any Huawei settlement, potential upside if deal closes. All other major OEM extensions are signed or pending execution.

Palkhiwala said Q4 margins beat the guide-down on stronger mix and Q1 is largely in line with Q4, a reasonable way to model the business going forward.

Brice pointed to display growth and ramp costs from adding customer-service engineers, but said margins are flattish at the company level near term.

Brice said slow improvement is the right framing and hiring continues a few quarters, with the headwind receding as revenue grows, but avoided a specific pace.

Brice confirmed the H2 math and noted more than 10 new factory projects added in the last quarter. Gary added 2027 still looks like a strong growth year driven by expanding compute demand.

Brice confirmed floor space is ready but needs fitting and hiring. Rough approximation of doubling capacity available, constrained by supply chain readiness.

Brice confirmed the math is in the ballpark and H2 weighting but declined to fill in quarterly specifics, saying they're still working out the schedule.

Brice confirmed cleanroom space is the binding constraint, noted customers have exhausted most available space this year, and highlighted multiple factory projects coming online in 2027.

Brice indicated semi business flattish through Q3 (their fiscal) with significant uplift in Q4/Q1, especially from leading edge. AGS will grow continuously at low double digits. Gross margin at 48.4% until volumes ramp in H2 calendar 2026.

Brice said Q3 China played out as expected. Leading-edge was slightly less than anticipated, gate-all-around now tracking ~$4.5B vs prior ~$5B expectation for fiscal 2025. Guided Q4 down ~$700M: ~$500M China + ~$500M leading edge, partially offset by rest-of-world ICAPS.

Brice acknowledged 200mm equipment was weaker than expected and utilization stayed flat, but core AGS hit a record and will grow low double digits. About 90% of AGS is now recurring revenue.

Brice confirmed year-over-year segment trends: leading logic accelerating, mature logic down, DRAM continuing strong, NAND ticking up. Sequentially, DRAM down, NAND could be up, foundry/logic up.

Brice clarified the China trade impact is front-loaded in Q2 and then sustained at that level, not incrementally worse. Core AGS grows low double digits stripping out 200mm equipment.

Brice confirmed underlying gross margin baseline improved to 48.0% through logistics, inventory management, cost reduction, and value pricing. Q1 guided 48.4% on favorable mix.

Management confirmed China normalized to 30% in Q4 and Q1 outlook, with ICAPS strong globally but watching automotive, industrial, and analog end markets.

Hu clarified acceleration refers to second half versus first half year-over-year growth, not sequential deceleration from Q2's 107%.

Stacy pressed on whether DC AI grew ex-China and Jean deflected, stating China revenue in Q1 was not material. Confirmed both DC AI and server will grow double digits sequentially in Q2.

Jean attributed OpEx overruns to aggressive AI investment driving revenue momentum and revenue beats pulling costs higher. Did not provide a specific full-year OpEx framework.

Jean Hu said the Q4 reserve release covered both Q4 and Q1 MI308 shipments, making the Q1 55% gross margin guide a clean number; Lisa Su declined to disclose annual Instinct revenue.

Lisa Su said 2025 OpEx lean-in was the right call and 2026 will deliver leverage with OpEx growing slower than revenue, especially as 2H revenue inflection hits.

Jean Hu said both data center sub-segments grew similarly year-over-year in Q3 with server slightly ahead; Lisa Su declined to give specific numbers for either component.

Lisa Su clarified the sequential Q4 data center guide as double-digit with both server and AI segments growing, noting the 'strong double-digit' language was applying to year-over-year commentary.

Lisa Su explicitly said Q2 DC GPU will not grow year-over-year due to the $700M MI308 removal, but Q3 and Q4 will grow year-over-year to deliver full-year strong double-digit growth.

Jean Hu confirmed DC segment down 5% sequentially with server slightly better and DC GPU down; declined to validate Stacy's specific magnitude estimates.

Lisa Su committed unambiguously that the 2025 GPU exit rate will exceed the 2024 exit rate, framing it as a prerequisite for strong double-digit full-year growth.

Lisa Su corrected Stacy's math, clarifying DC segment was up 9% sequentially with server slightly above and GPU slightly below; said first half 2025 should be roughly flat with second half 2024.

Gave a concrete fab count as the capacity gate but declined to convert it into a 2027 WFE number.

Signalled sequential confidence without committing, explicitly deferring a quarter-by-quarter framework to a later call.

Doug clarified the $40B includes both upgrades and new equipment purchases for bottleneck-breaking, with upgrades already quite strong in 2025 and continuing.

Doug declined to specify 2027 WFE magnitude but confirmed clean rooms will be more available and WFE will grow nicely; Tim added rising etch/dep intensity benefits Lam.

Doug said he expects sequential growth every quarter through 2026, with second-half weighting from both WFE and revenue perspectives, modulated by clean room space availability.

Doug confirmed Stacy's directional estimate of ~$1.5B China headwind and said global multinationals will more than offset it, driven by NAND, HBM, and advanced nodes.

Doug attributed December softness to normal profile, less favorable mix, and incrementally higher tariffs, not unusual or conservative.

Doug acknowledged gross margin variability around the current 49% level depending on customer mix, product mix, and revenue levels, without committing to a direction.

Doug explained Reliant declines from losing a bunch of China customers are the main drag offsetting upgrade strength in the CSBG flattish outlook.

Management said Lam captures a higher share of spend and outperforms more in an upgrade year, though the industry spends less overall, declining to give numbers.

Haviv expects PE to grow sequentially but at a lower level, citing shortages pressuring customers; declined a specific by-end-market guide.

Rafael guided OpEx, net OIE and acquisition charges to be flat sequentially to aid gross-margin modeling.

Rafael identified the missing piece as Silicon Labs acquisition charges that will continue every quarter until close. Confirmed OpEx assumption was reasonable.

Rafael confirmed TI plans to continue GAAP reporting post-acquisition, providing components for analysts to construct their own non-GAAP views. Referenced NAT Semi precedent.

Haviv confirmed Q4 is roughly seasonal reflecting moderate recovery pace. Mike noted Q1 historically is typically down slightly sequentially but declined to guide.

Rafael confirmed Q4 GPM will decline from lower revenue, higher depreciation, and reduced loadings, but stepped back to emphasize trailing 12-month free cash flow up 65%.

Rafael corrected that GPM will actually be above flat for Q3 despite higher depreciation, with the miss in models likely from unfavorable net OI&E of ~$20M from lower cash/higher debt interest.

Haviv explained 4 of 5 markets now recovering but auto still lagging, and the tariff/geopolitical environment requires prepared flexibility despite strong cyclical signals.

Haviv clarified Q1 was genuine cyclical recovery (broad, not anxious), but Q2 may include some precautionary inventory build. Committed to disciplined demand management playbook.

Mike Beckman said Q2 linearity so far is about as expected for a recovery environment, without isolating a post-tariff acceleration. Difficult to attribute specific orders to tariff anxiety.