Tan's First Quarter: A Beat Helped by Some Tariff Pull-Forward, A Guide That Pulls the Rug
Key Takeaways
- Revenue of $12.7B beat consensus by ~$370M and landed at the high end of guidance — but management said it believes Q1 revenue benefited from customers buying ahead of potential tariffs (a magnitude it called difficult to quantify), meaning part of the beat is likely borrowed from Q2 rather than earned.
- Q2 revenue guide of $11.2–12.4B (midpoint $11.8B) sits ~$1B below Street — the widest quarterly miss-vs.-consensus on guidance in recent memory, and CFO Zinsner explicitly flagged rising recession probability as the driver of the wide range.
- New CEO Lip-Bu Tan, in his first earnings call since taking the role in March, prioritized organizational flattening and cultural change over specific financial targets — OpEx cut to $17B (from $17.5B) with a $16B bogey for 2026, but layoffs were not sized: Zinsner said the headcount impact had not yet been identified.
- Altera divestiture (51% to Silver Lake at ~$9B EV, ~$4.4B net cash) confirms the "focus on the core" thesis, but negative adjusted FCF of $3.7B this quarter underscores that foundry cash burn remains the dominant near-term valuation question.
- Rating: Initiating at Hold. A first-time CEO, a guide cut, and an unquantified restructuring create too many moving parts to underwrite either direction with conviction at current levels — we want to see one clean quarter of execution before upgrading, and one clear 18A slip before downgrading.
INTC financial model
Income Statement · Dollars in millions, except per share
| Income Statement | ||||||
|---|---|---|---|---|---|---|
| Actual | Estimate | |||||
| FY2023 | FY2024 | FY2025 | FY26E | FY27E | FY28E | |
| Revenue | $54,228.0 | $53,101.0 | $52,853.0 | $56,276.3 | $61,536.9 | $70,597.7 |
| Less: Cost of Sales | ($32,517.0) | ($35,756.0) | ($34,478.0) | ($33,123.7) | ($32,898.6) | ($36,710.8) |
| Gross Profit | $21,711.0 | $17,345.0 | $18,375.0 | $23,152.6 | $28,638.3 | $33,886.9 |
| Less: Research and Development | ($16,046.0) | ($16,546.0) | ($13,774.0) | ($14,040.2) | ($14,300.5) | ($15,531.5) |
| Less: Marketing, General and Administrative | (5,634.0) | (5,507.0) | (4,624.0) | (4,667.4) | (4,996.4) | (5,506.6) |
| Less: Restructuring and Other Charges | 62.0 | (6,970.0) | (2,191.0) | (4,320.0) | 0.0 | 0.0 |
| Total Operating Expenses | ($21,618.0) | ($29,023.0) | ($20,589.0) | ($23,027.6) | ($19,296.9) | ($21,038.1) |
| Operating Income (Loss) | $93.0 | ($11,678.0) | ($2,214.0) | $125.0 | $9,341.4 | $12,848.8 |
| Gains (Losses) on Equity Investments, Net | $40.0 | $242.0 | $514.0 | $78.0 | $200.0 | $200.0 |
| Interest and Other, Net | 629.0 | 226.0 | 3,257.0 | (588.0) | 200.0 | 200.0 |
| Income (Loss) Before Taxes | $762.0 | ($11,210.0) | $1,557.0 | ($385.0) | $9,741.4 | $13,248.8 |
| Less: Provision for (Benefit from) Taxes | $913.0 | ($8,023.0) | ($1,531.0) | ($1,082.8) | ($2,045.7) | ($2,782.2) |
| Net Income (Loss) | $1,675.0 | ($19,233.0) | $26.0 | ($1,467.8) | $7,695.7 | $10,466.5 |
| Less: Net Income (Loss) Attributable to NCI | ($14.0) | ($477.0) | $293.0 | ($703.0) | ($200.0) | ($200.0) |
| Net Income (Loss) Attributable to Intel | $1,689.0 | ($18,756.0) | ($267.0) | ($764.8) | $7,895.7 | $10,666.5 |
| EPS, Basic | $0.00 | $0.00 | $0.00 | ($0.15) | $1.52 | $2.02 |
| EPS, Diluted | $0.40 | ($4.38) | ($0.06) | ($0.15) | $1.52 | $2.02 |
| Weighted Avg Shares — Basic (M) | 0.000 | 0.000 | 0.000 | 5,128.000 | 5,200.000 | 5,275.000 |
| Weighted Avg Shares — Diluted (M) | 0.000 | 0.000 | 0.000 | 5,128.000 | 5,200.000 | 5,275.000 |
| Ratios & Assumptions | ||||||
| YoY Revenue Growth % | (14.0%) | (2.1%) | (0.5%) | 6.5% | 9.3% | 14.7% |
| Gross Margin % (driver) | 40.0% | 32.7% | 34.8% | 41.1% | 46.5% | 48.0% |
| Operating Margin % | 0.2% | (22.0%) | (4.2%) | 0.2% | 15.2% | 18.2% |
| Net Margin % (consolidated) | 3.1% | (36.2%) | 0.0% | (2.6%) | 12.5% | 14.8% |
| Effective Tax Rate (driver) | (119.8%) | (71.6%) | 98.3% | (281.3%) | 21.0% | 21.0% |
The full workbook adds 16 historical and 8 projected quarters, plus KPI Drivers · Balance Sheet · Cash Flow Statement — every subtotal a live formula, every projection traced to a driver.
Results vs. Consensus
| Metric | Actual | Consensus | Beat/Miss | Magnitude |
|---|---|---|---|---|
| Revenue | $12.67B | $12.30B | Beat | +3.0% |
| Non-GAAP Gross Margin | 39.2% | ~36.0% | Beat | +320 bps |
| Non-GAAP Operating Income | $690M | ~$380M | Beat | +82% |
| GAAP EPS | $(0.19) | $(0.23) | Beat | +$0.04 |
| Non-GAAP EPS | $0.13 | $0.01 | Beat | +$0.12 |
| Adjusted FCF | $(3.68)B | n/a | Negative | — |
Quality of Beat/Miss
- Revenue: Beat was low-quality. Zinsner said "we believe Q1 revenue benefited from customer purchasing behavior" ahead of potential tariffs, while calling the magnitude difficult to quantify, and the Q2 guide compensates by cutting deep. Strip out the pull-forward and the underlying run-rate likely looks closer to the low end of prior guidance.
- Margins: 320 bps non-GAAP GM beat was the highlight of the print — attributed to the higher revenue and a mix shift toward Raptor Lake versus Lunar Lake, plus improved Meteor Lake cost. Zinsner said that without the revenue beat and the Raptor Lake mix, margin "would've been much closer to the guide." Neither driver is durable: he expects client mix to move toward Lunar Lake through the year, a margin headwind.
- EPS: Operational, not engineered. Limited below-the-line help; Zinsner credited higher revenue, stronger gross margin and lower operating expenses.
- FCF: $(3.7)B adjusted FCF is the number that matters for the bear case. Capex ran hot against still-depressed operating cash flow; OpEx and capex cuts announced today are real, but the Q1 cash profile is the baseline the Street will extrapolate until proven otherwise.
Segment Performance
| Segment | Revenue | YoY Growth | vs. Estimate | Notable |
|---|---|---|---|---|
| Client Computing (CCG) | $7.6B | -8% | In line to slight beat | Mix shifted back toward older Raptor Lake on price sensitivity |
| Data Center & AI (DCAI) | $4.1B | +8% | Beat | Xeon strength drove the topline beat |
| Intel Foundry | $4.7B | +7% | In line | Still loss-making; internal demand dominates |
| All Other (Altera, Mobileye, etc.) | $0.9B | +47% | Beat | Altera recovery; to be deconsolidated post-Silver Lake close |
Client Computing (CCG)
CCG at $7.6B down 8% YoY was the soft spot and the one that most complicates the Q2 setup. Unit volumes reportedly held up — but mix shifted back toward older Raptor Lake SKUs rather than Meteor Lake/Lunar Lake, which carry higher ASPs. This is uncomfortable: it says the premium AI-PC story Intel has been telling for 18 months is not yet translating into preferential mix.
"Raptor Lake is a great part. Meteor Lake and Lunar Lake are great as well, but come with a much higher cost structure, not only for us, but at the system, ASP price points for our OEMs as well." — Michelle Johnston Holthaus, CEO Intel Products
Assessment: Whether this is cyclical or structural is the $20B question. Our read: it's mostly cyclical, but the Q2 guide implies no immediate mix recovery, and that keeps CCG gross-margin progress capped through at least the summer.
Data Center and AI (DCAI)
DCAI at $4.1B up 8% YoY was the strongest segment. Zinsner attributed the upside to hyperscaler demand for host CPUs for AI servers and storage compute. This is the first quarter in several where DCAI growth wasn't overshadowed by a hyperscaler inventory digestion excuse, though Holthaus said a few hyperscalers drove Q1, allowed that it could be read as "hedging your bets or doing pull-ins," and does not expect it to continue into Q2.
Assessment: Positive, but context-dependent. DCAI growth is being measured against a depressed FY2024 base, and the competitive setup vs. AMD EPYC and custom ASIC silicon is no easier than it was 90 days ago. One quarter does not make a trend.
Intel Foundry
IFS revenue of $4.7B up 7% is almost entirely internal Intel Products volume. External foundry traction remains slow; Tan put ramping the internal customer first, with building trust with external foundry customers as the next step and yield and reliability as the process priorities.
"It is first and foremost a customer service business built on the foundational principle of trust. And we need to instill a customer service mindset across our foundry business." — Lip-Bu Tan, CEO
Assessment: The strategic message shifted meaningfully. The prior regime treated IFS as a revenue growth engine; Tan is treating it as an operations problem first. That's probably the right diagnosis — but it also implies foundry financials get worse before better, because the discipline will come before the demand.
Guidance & Outlook
| Metric | Q2 Guide (Low) | Q2 Guide (High) | Q2 Guide (Mid) | Consensus | Delta vs. Street |
|---|---|---|---|---|---|
| Revenue | $11.2B | $12.4B | $11.8B | $12.82B | -8.0% |
| Non-GAAP Gross Margin | — | — | 36.5% | 36.9% | In line |
| Non-GAAP EPS | — | — | $0.00 | $0.06 | Miss |
Guide is ~$1B light at the midpoint, and the range is unusually wide ($1.2B band). Both signal management discomfort with visibility, not just caution.
"The very fluid trade policies in the US and beyond as well as regulatory risks have increased the chance of an economic slowdown with the probability of a recession growing." — David Zinsner, CFO
Guidance style: Notably more conservative than the prior regime's pattern. Tan/Zinsner appear to be setting a low bar they can clear, which is a common new-CEO playbook — but it also means the Q2 print needs to clear the low bar cleanly to rebuild credibility.
Street at: Consensus for Q2 will reset down to roughly $11.8–12.0B post-guide; full-year 2025 estimates likely trimmed 4–6%.
Key Topics & Management Commentary
Overall Management Tone: New-CEO candor, to the point of discomfort. Tan's prepared remarks were unusually direct about Intel's structural problems — "bureaucracies have been suffocating the innovation and agility we need to win" is not a line the prior regime would have allowed. Zinsner's macro language was similarly blunt. The pitch is: we see the problems, we own them, we're going to fix them, and we will not promise specifics we cannot deliver.
Leadership Reset
Tan used his first call to signal a flatter org structure, with all critical product, manufacturing, and G&A functions now reporting directly to him. A four-day return-to-office mandate takes effect in Q3. Headcount cuts were implied but not sized: Zinsner said Intel had "not yet identified what that means from a headcount perspective" and promised more detail on the July call.
"Organizational complexity and bureaucracies have been suffocating the innovation and agility we need to win. It takes too long for decisions to get made." — Lip-Bu Tan, CEO
Assessment: This is the right first move — every large-cap turnaround starts with span-of-control and accountability fixes. But the absence of a layoff number creates a dangling overhang: headlines will trickle out over the next 60–90 days, each one re-raising the question, and the stock will be reactive to each leak. That's a volatility tax Intel could have paid in one lump.
Foundry Strategy Reset: Quality Before Scale
Perhaps the most meaningful strategic shift: Tan put the internal ramp ahead of new external foundry customers. Priority one is ramping the internal customer (Panther Lake on Intel 18A), with yield and reliability as the process focus; building trust with external foundry customers is the next step. Panther Lake first SKU remains on track for year-end 2025, with additional SKUs in H1 2026, and Holthaus said the goal of an approximately 70% in-house wafer mix "remains steadfast," with Panther Lake helping Intel get there.
"I think the priority one for us is for Intel Foundry is ramping our internal customer. … And then the next step is basically building the trust with our foundry customer and, you know, the PDK, the schedule, you know, to really make sure that we're really robust on that." — Lip-Bu Tan, CEO, answering Timothy Arcuri (UBS)
Assessment: Strategically sound, financially painful. IFS will keep losing money in 2025 — possibly through 2026 — while the operational work gets done. This positioning effectively removes "external foundry revenue" as a near-term bull catalyst.
AI Strategy — Reframed Toward Inference and Edge
Tan's AI pitch de-emphasized the training GPU race (where Intel is a distant third behind Nvidia and AMD) and leaned into reasoning models, edge inference, and enterprise deployment — areas where the Xeon + Gaudi + AI-PC stack has more natural right-to-win.
Assessment: This is a more honest framing of Intel's AI position than the prior regime offered. The challenge: reasoning-model inference is a real market, but it's not where most of today's AI capex is going, and the ROI case is longer dated than hyperscaler training spend.
Altera Sale and Portfolio Focus
Intel is selling 51% of Altera to Silver Lake at a ~$9B enterprise value, yielding ~$4.4B in net cash proceeds. The deal deconsolidates Altera from Intel's P&L going forward, with OpEx coming down dollar-for-dollar as Altera exits.
Assessment: Good execution at a reasonable price. This is a cleaner outcome than the 2023 Altera IPO plan and frees up management bandwidth. The $4.4B doesn't move the needle on capex needs, but every billion counts when FCF is negative.
Tariff Pull-Forward: The Awkward Footnote
Management said it believes part of Q1 revenue reflected customers buying ahead of potential tariffs, though Zinsner said the magnitude is difficult to quantify. That helps explain why Q2 guide is so much weaker than Q1 results. But it also means the "beat" is a borrow against Q2 — and if tariffs actually materialize, Q3/Q4 get dragged into the mess as well.
Assessment: The pull-forward admission is helpful for credibility but damaging for run-rate. We assume ~$300–500M of Q1 revenue was pulled forward; that alone explains roughly a third to half of the ~$1B Q2 shortfall vs. Street.
Analyst Q&A Highlights
Turnaround Timeline
- Vivek Arya, Bank of America: Asked how long investors should realistically expect the turnaround to take (one, two or three years) and which metric Intel is optimizing for: share gains, gross margin or free cash flow. Tan gave no calendar, repeating "there's no quick fix," and named product metrics instead: the best products, power efficiency, performance and schedule ("time to market is critical on schedule").
Assessment: Refusing to anchor a timeline is consistent with new-CEO posture. But it also leaves the Street to extrapolate, which usually means they extrapolate pessimistically.
Product Mix and Raptor Lake Demand
- Stacy Rasgon, Bernstein: Asked why demand is shifting to older Raptor Lake over newer Meteor Lake/Lunar Lake, and whether Intel is "pushing the older gen stuff" because the margins are better. Holthaus rejected the margin reading ("We're not pushing the old parts"), attributing the shift to customers protecting system price points while macro and tariff concerns have everyone "hedging their bets," and to the much higher cost structure of Meteor Lake and Lunar Lake for Intel and its OEMs.
Assessment: The answer is internally consistent, but it doesn't explain why Intel didn't reprice Meteor/Lunar more aggressively to defend mix. That's either a margin choice or an OEM negotiation gap — neither framing is comforting.
IDM Model Viability
- Vivek Arya, Bank of America: Asked directly whether the integrated device manufacturer (IDM) structure still makes sense given negative foundry margins. Tan offered a "balanced approach" using TSMC as a partner while driving foundry efficiency — notably, he did NOT offer a full-throated defense of the combined Products + Foundry structure.
- Notable dodge: Tan never said "we are committed to the IDM model." That omission is doing a lot of work and will be parsed heavily by the Street.
Capital Intensity and Foundry Cash Burn
- Vijay Rakesh, Mizuho: Asked how Intel would approach capex across 2025–26 and whether it is rationalizing the fab footprint. Zinsner said Intel can operate to $18B in 2025, citing "$50 billion or so of assets under construction," much of it "equipment that's still in bubble wrap," called it too early to guide 2026 capex, and named roughly 25% capital intensity under a full IDM model as a good long-term working assumption. The $8–11B net capex range (net of government incentives and partner contributions) came in the prepared remarks.
Assessment: The call never revisited a foundry breakeven date, and no analyst asked. Prior management pointed to 2027 breakeven; Tan/Zinsner did not reaffirm it, and we read that silence as the most important signal of the call.
What They're NOT Saying
- Layoff sizing: Analysts' questions assumed headcount reductions and Zinsner flagged likely restructuring charges, but he said Intel had not yet identified the headcount impact and promised more detail in July. This will leak to the press in pieces over the next 60–90 days and each leak will be a volatility event. The open-ended answer is unusual and suggests the number may be larger than consensus expects (>15% of workforce).
- Foundry breakeven year: Prior regime had pointed to 2027 IFS operating breakeven. Tan/Zinsner conspicuously did not reaffirm that target. The implication: 2027 is off the table, but no new date has been set.
- Gaudi / AI accelerator roadmap: One analyst question and no accelerator detail in the prepared remarks. Holthaus restated that Falcon Shores will not be built and Jaguar Shores stays on the roadmap, said the AI roadmap will be discussed externally "over the coming months," and conceded it is a segment "we're not competing in today." No Gaudi revenue, customer or timing detail was offered. The AI accelerator strategy is being reset; what replaces it is still undefined.
- IDM commitment: Tan declined to explicitly endorse the IDM model. He did not float a split either — but the silence itself is informative and will drive significant "breakup speculation" noise.
- CHIPS Act milestone detail: Zinsner cited $1.1B received from CHIPS grants and held net capex at $8–11B "due to uncertainty regarding the timing of the US government fulfilling their obligations in our CHIPS agreement," but gave no milestone-level detail. Given the change in Washington, that thin disclosure is notable.
Market Reaction
- After-hours move (April 24): From a $21.49 pre-print close, an initial small pop on the beat faded to -6 to -8% as the Q2 guide hit the tape.
- Next-day session (April 25): Not yet traded at this note's 8:00 AM ET publication; the after-hours move is the latest price signal.
- Setup entering the print: +7.2% YTD (S&P 500 -6.7%), -11.2% over the trailing 30 days and -37.7% over 12 months; 52-week closing range $18.13–$35.11.
- Analyst reactions (by publication):
- Sell-side notes were still arriving at this note's 8:00 AM ET publication; no target changes are cited here.
- Tone shifted from new-CEO optimism to a show-me story.
- No material upgrades.
The after-hours sell-off is almost entirely a guidance story, not a quarter story. If Q2 had been guided to $12.5–12.8B, the stock likely would have printed flat-to-up on the Q1 beat and the credible new-CEO framing. The $1B guide cut, the unsized layoffs, and the $(3.7)B adjusted FCF together made the print impossible to buy on the news.
Street Perspective
Debate: Is the New-CEO Setup Buyable Now or Does It Need Proof?
Bull view: Tan inherits a de-rated stock, a lowered bar, and a clean slate to reset expectations. Execution against a low bar is the highest-probability path to multi-quarter outperformance; patient investors should be adding under $20.
Bear view: Intel has been a "next quarter it turns" story for three years running. The new-CEO narrative is narratively attractive but financially unsupported — FCF is deeply negative, foundry has no demand, and client mix is degrading. Buying ahead of proof is a behavioral bias, not a thesis.
Our take: The bears have the stronger argument today. Tan's strategic diagnosis is correct, but Q2 guide tells us execution will get worse before better. We'd rather pay 15–20% more after one clean print than catch a falling knife before it.
Debate: Does the IDM Model Survive?
Bull view: Tan's decision to fix foundry operations before pursuing new external customers implies the IDM model is being preserved and strengthened; a split would have been announced today if it were on the table.
Bear view: Tan's refusal to explicitly endorse the IDM model, combined with the Altera divestiture and the silence on the 2027 foundry breakeven, suggests the combined structure is under active review and could be split within 12–18 months.
Our take: We think the IDM model survives, but in a meaningfully thinner form — more TSMC reliance, more external capacity, less fab-first identity. That's still structurally different from a full split, but it's not the prior regime's IDM 2.0 either.
Debate: Is 18A Actually on Track?
Bull view: Panther Lake Q4 2025 launch was reaffirmed; this is the same date Intel has held for 18 months, which implies yield and schedule progress has been real even if uncelebrated.
Bear view: Every TSMC and Samsung node transition this cycle has slipped at least once; assuming Intel 18A alone escapes that pattern is betting on an outlier. And if it slips, the knock-on to Panther Lake financials and the entire foundry credibility story is severe.
Our take: A 1–2 quarter slip on 18A is the single biggest risk to the bull case over the next 12 months, and today's call did nothing to reduce that risk. Management didn't offer the kind of specific yield/defect-density metrics that would let analysts get comfortable.
Model Update Needed
| Item | Prior Street Model | Suggested Change | Reason |
|---|---|---|---|
| FY25 Revenue | ~$54B | $51–52B | Q2 guide $1B light; H2 unchanged but the Q1 pull-forward is a debit |
| FY25 Non-GAAP GM | ~37.0% | 36.8% | Q1 beat offsets some of the H2 mix pressure |
| FY25 OpEx | ~$17.5B | $17.0B | Explicit management cut |
| FY25 Non-GAAP EPS | ~$0.80 | $0.55–0.65 | Revenue and mix cuts flow through |
| FY25 CapEx | ~$20B gross | $18B gross | Explicit management cut |
| FY25 Adjusted FCF | ~$(4)B | $(6)–(8)B | Q1 set a worse baseline than Street expected |
Valuation impact: At the $21.49 pre-print close (6–8% lower after hours), INTC trades just under 1x book ($22.87 per share at quarter-end) and ~33–39x our $0.55–0.65 trough FY25 non-GAAP EPS estimate. The earnings multiple is meaningless in a turnaround year; book-value and replacement-value frameworks dominate. We'd anchor to ~$19–23 fair value over the next two quarters, expanding to $22–28 only when a clean IFS milestone or a layoff sizing event removes one of the overhangs.
Thesis Scorecard Post-Earnings
| Thesis Point | Status | Notes |
|---|---|---|
| Bull #1: New CEO will unlock operational improvement | Neutral | Right strategic diagnosis; but no specifics on timeline or layoff sizing |
| Bull #2: 18A node ramp restores manufacturing leadership | Neutral | Schedule reaffirmed but not proven; no yield metrics shared |
| Bull #3: DCAI Xeon refresh returns to share gains | Partly supported | DCAI +8% YoY, but Holthaus tied Q1 strength to a few hyperscalers (possible pull-ins) and does not expect it to continue into Q2 |
| Bear #1: Foundry is a structural cash drain with no customers | Confirmed | Tan put the internal ramp ahead of new external customers; 2027 breakeven not reaffirmed |
| Bear #2: Client PC mix is degrading toward older, lower-ASP SKUs | Confirmed | Raptor Lake outselling Meteor/Lunar Lake is a material negative signal |
| Bear #3: Macro/tariff risk compounds an already weak setup | Confirmed | Zinsner explicitly flagged recession risk in prepared remarks |
Overall: Thesis neutral-to-weakened. Bull points require time and proof; bear points got multiple confirming data points today.
Action: Initiate at Hold. Wait for one clean quarter of Q2 2025 execution before upgrading. Downgrade to Underperform on any 18A slip, a worse-than-expected layoff sizing (>20% of workforce), or a Q3 guide that fails to resume sequential growth.