A Blowout Print the Stock Had Already Bought: Downgrading to Hold as the Seventh Straight Beat Fades to a Loss

Results vs. Consensus

Year-over-Year Comparison (Q2 2026 vs. Q2 2025)

Sequential Comparison (Q2 2026 vs. Q1 2026)

Revenue

The $16.1B print is the cleanest revenue beat Intel has produced in years, and its composition matters more than its size. Growth was broad, led by the two AI-exposed engines: DCAI up 59% and CCPG up 13% year-over-year, with Intel Foundry up 31%. Collectively, management said the "AI-driven businesses grew greater than 70% year-over-year" and contributed roughly 70% of revenue. Crucially, the beat was not a demand pull-forward of the kind that flattered the Q1 2025 print; it was Intel finally converting more of its backlog into shipments as factory output improved. CFO Zinsner was explicit that even after exceeding wafer-out targets, "strengthening demand continues to outstrip our growing supply." A supply-limited revenue beat is the highest-quality kind, because the ceiling is Intel's own execution rather than the market's appetite.

Margins

Non-GAAP gross margin of 41.8% is the second consecutive quarter comfortably in the 40s, and Zinsner made clear that clearing that bar in every quarter was the explicit finance-organization objective for the year. The drivers are the right ones: volume leverage, better 18A yields, and ASP gains from mix and pricing. There is a genuine offset building underneath, however. Panther Lake and Granite Rapids are now large enough in the mix that their below-corporate-average early-ramp margins are a drag, which is why the Q3 guide holds gross margin roughly flat at 42.0% despite the loss of a Q2 client inventory write-down that would otherwise have been a tailwind. The trajectory is up, but the near-term slope is gentle, not steep.

EPS and the GAAP Loss

Non-GAAP EPS of $0.42 is operating quality: a swing from a $(0.10) loss a year ago, driven by the $3.3B year-over-year improvement in non-GAAP operating income, not by tax or share count. The $(2.16) GAAP loss is a different animal and demands care. It is fully explained by a single non-cash item: a roughly $12.5B mark-to-market charge on the Escrowed Shares derivative liability granted to the US government under the CHIPS Act Secure Enclave agreement. That liability marks up as Intel's stock rises, so the better the shares perform, the larger the GAAP loss. INTC's move from the mid-$30s to ~$100 over the trailing year is what produced an $11.0B GAAP net loss in a quarter of record operating strength. It is accounting, not deterioration, and models should adjust it out. It is also, however, a live reminder that the government stake carries an ongoing, counterintuitive P&L cost that will keep GAAP results noisy.

Segment Performance

Data Center and AI (DCAI) — The Engine Reaches Full Power

DCAI is the quarter's centerpiece and the single most convincing segment print of the turnaround. Revenue of $6.26B grew 59% year-over-year and 24% sequentially, and it did so profitably: operating income of $2.47B at a 39.5% margin, up roughly $1B sequentially and nearly quadrupling from $633M a year ago. Growth was units-led with rising core counts, which is the highest-quality composition available; on a market this large and this capacity-constrained, unit growth with core-count mix implies both socket-share resilience and ASP uplift. Xeon 6 remains one of the fastest-ramping products in Intel's history, and the segment launched Xeon 6+ (Clearwater Forest), its first server-class product on 18A. Purpose-built silicon inside DCAI grew about 20% sequentially and nearly tripled year-over-year.

"As AI expands from training to inference and increasingly to agentic and multi-agent systems, general purpose server CPU density continue to increase, and our core server CPU franchise is growing faster than ever." — Lip-Bu Tan, CEO

Assessment: The bear case that framed Intel as collateral damage in the AI capex shift is not just refuted, it is running in reverse. A 59% year-over-year increase at a 40% operating margin is structural reacceleration, and management's framing that the CPU-to-GPU ratio is approaching parity gives the segment a multi-year demand narrative. The one caveat is durability: this is a supply-constrained peak, and part of the extraordinary margin reflects operating leverage on a volume the company itself says it cannot yet fully serve. It is excellent, and it is being priced as if it is permanent.

Client Computing and Physical AI (CCPG) — Renamed, Repriced, and Quietly Strong

The renamed client segment (formerly CCG) delivered $8.88B, up 13% year-over-year and 15% sequentially, ahead of plan. The upside, however, was largely ASP rather than units: Zinsner attributed it to mix skew toward the high end and like-for-like price increases passing through cost inflation, against a PC end-market that is actually softening. AI PC revenue grew 26% sequentially and is now two-thirds of the client mix, and the edge and physical-AI push (about 10% of CCPG revenue) is the reason for the rename. Operating income of $2.34B (26.4% margin) was down about $173M sequentially on an inventory write-down for stranded, unmatched-set product.

"During a period of inflationary pressures, CCPG brought Core Series 3 to market at the ideal time on its A stepping to provide cost-optimized mainstream compute capabilities." — David Zinsner, CFO

Assessment: CCPG is doing its job, funding the rest of the company while the market it serves weakens. Management expects PC consumption to be subseasonal in the second half and down low double digits for the full year, and candidly plans to pivot wafers away from client toward the far-more-constrained server business in Q4. That is rational allocation, but it means client is a source of cash and wafers, not a growth vector. The ASP-driven nature of the beat also makes it lower-quality than the DCAI print: pricing through inflation is not the same as winning units.

Intel Foundry — The P&L Improves; the Business Case Still Waits

Foundry revenue of $5.77B rose 31% year-over-year and 6% sequentially on higher fab volumes, with 18A output roughly 25% above target and up more than 50% sequentially. The operating loss narrowed to $(2.09B), a $348M sequential and ~$1.1B year-over-year improvement, as better yields and factory scale drove down wafer costs; the primary Panther Lake SKU cost is down about 50% year-to-date with another 20% targeted this year. 18A-P entered risk production. Yet the number that matters most for the thesis, external foundry revenue, was just $293M, and the named external anchor that defined last quarter's story went unmentioned on this call.

"The successful volume ramp of 18A for our internal products provide important validations as our Intel Foundry engaged with external customers." — Lip-Bu Tan, CEO

Assessment: Foundry is executing well as a cost-reduction story and is still pre-commercial as a merchant-foundry story. The internal 18A ramp is real and is the proof point external customers want to see, but "engaged with external customers" and $293M of external revenue is a long way from the third-party-demand inflection the equity is being asked to underwrite. The decision to commit 14A to high-volume manufacturing in 2028 (see Key Topics) is the most important forward signal here, and it is being made substantially on the strength of internal demand plus engagement, not on a book of signed external commitments.

Key Topics & Management Commentary

Overall Management Tone: Confident, disciplined, and notably more expansive on capital deployment than at any point in the Tan era. Management leaned into the supply-shortage framing as an opportunity rather than a constraint, and for the first time paired it with a materially higher capex commitment. The posture on customers was unchanged from prior quarters: describe the engagement, decline to name the customer, and let the internal-execution track record carry the credibility. Where the call was less forthcoming was precisely where last quarter had been most forthcoming, on named external foundry demand, and the silence on the prior quarter's marquee 14A relationship was conspicuous against how central it had been ninety days earlier.

Topic 1: The Sell-the-News — A Beat the Market Had Already Bought

The defining fact of this earnings event is not in the release; it is the stock's reaction to it. Intel cleared every guided metric, raised the forward outlook well above Street, and de-risked its roadmap, and the shares closed down 3.0%. The explanation is positioning and valuation, not results. Entering the print INTC had compounded roughly 327% over twelve months to a record near $141, corrected about 28% during July, and still traded near 90x forward non-GAAP earnings. Prediction markets reportedly assigned near-certainty to a beat. When expectations are that saturated, even a blowout leaves nothing to buy.

This is the mirror image of the January setup. After Q4 2025's guide miss the stock fell 17% into the low $30s, and we argued the sell-off created the entry point. The same discipline applied symmetrically now points the other way: a fully-priced stock that fades its best quarter is signaling that the reward for being right on the fundamentals has already been collected.

Assessment: A stock that cannot rally on a two-times EPS beat and an 8%-above-Street guide is telling you where it is in its cycle. The turnaround narrative that powered the re-rating from $19 to $100 is now consensus, and consensus, capitalized at 90x, is not where forward returns are made. This single dynamic is the basis for the rating change.

Topic 2: 14A Committed to 2028 High-Volume Manufacturing — Roadmap De-Risked, Anchor Gone Quiet

The most consequential forward disclosure was that Intel made the decision in Q2 to fully commit 14A to high-volume manufacturing in 2028, with risk production for internal products in the second half of 2027. Management reported that 14A defect density and transistor performance are outpacing 18A at the equivalent stage, that PDK 0.5 is complete, and that PDK 0.9 is on track for October. Committing capital to a 2028 HVM ramp is a real vote of confidence in the node's competitiveness and in the demand behind it.

"With encouraging external customer progress and increased demand for our internal products, we remain on track for 14A risk production for our internal products in second half of 2027, and we make the decision in Q2 to fully committed to high-volume ramp in 2028." — Lip-Bu Tan, CEO

What is missing is as notable as what was said. Last quarter's centerpiece, the Tesla/Terafab relationship that Intel presented as the named 14A anchor and that underpinned our "maximum conviction" framing, was not mentioned once on this call. The 2028 commitment was justified by internal-product demand and generalized "customer engagements," not by a reaffirmed external anchor. That does not mean the relationship has changed, but a named marquee customer that vanishes from the narrative one quarter after being introduced is a data point, and external foundry revenue of $293M is the quantitative version of the same caution.

Assessment: The node itself looks increasingly credible, and committing to 2028 HVM removes a binary that hung over the roadmap. But the investable question has quietly shifted from "will 14A land external anchors" back toward "when does external demand actually show up in revenue." The commitment is a positive; the silence on the anchor is a reason to hold conviction, not raise it.

Topic 3: Capex Raised to Over $20B — Capital Intensity Rises Into Negative Free Cash Flow

Management raised 2026 capital expenditure guidance to more than $20B, up roughly $3B versus the plan entering the year, and signaled 2027 will be "significantly above" 2026, with the vast majority spent across the US network. Tooling alone is up about 40% versus 2025. The rationale is demand-signal-driven, and Zinsner tied it explicitly to long-term agreements that now give multi-year visibility.

"Due to strong customer demand signals, we're raising our outlook for 2026 and now expect our CapEx to be more than $20 billion... we're forecasting 2027 capital expenditures to be significantly above the 2026 levels." — David Zinsner, CFO

The counterweight is cash. Adjusted free cash flow was $(8.4B) in the quarter, and while much of that reflects the one-time $12.2B net partner distribution tied to the Fab 34 buyback, the underlying direction of travel is a rising capex commitment against a business that does not yet generate free cash. Zinsner acknowledged that back-end and substrate investments "could be a drag" that makes reaching positive free cash flow next year "a little bit more challenged," and, tellingly, allowed that "if we're super successful... we may need to tap the capital markets."

Assessment: Rising capex on genuine demand is defensible, and Intel's discipline of putting capital in only behind confirmed demand is the correct posture. But for an equity trading at 90x forward, the combination of increasing capital intensity, negative free cash flow, and an explicit acknowledgment that more external capital may be required is exactly the kind of profile that limits multiple expansion from here. The P&L is inflecting faster than the cash flow, and valuation this full is usually paid for with cash, not accounting earnings.

Topic 4: 18A in Volume — Panther Lake and the Cost Curve

18A is now in high-volume manufacturing across multiple client and commercial products, including a subset of Panther Lake built using ASML's EXE High-NA EUV tools, with output increasing sequentially every month and yields tracking ahead of expectations. The economic story is the cost curve: the primary Panther Lake SKU is down about 50% year-to-date with another 20% targeted this year, which is the mechanism by which client margins move from a drag to a tailwind over the next several quarters. Clearwater Forest (Xeon 6+) extended 18A into the server line.

"18A output increased meaningfully in the quarter. Yields continue to track ahead of expectations. We are now ramping multiple new products on 18A, while supporting growing demand for our lead products, including Panther Lake and Wildcat Lake." — Lip-Bu Tan, CEO

Assessment: 18A execution is the load-bearing pillar of the entire thesis and it continues to hold. The yield and cost trajectory is what eventually lifts Panther Lake above corporate-average margin and validates Intel's leading-edge manufacturing claim to external customers. This pillar is on track and remains the strongest part of the fundamental case.

Topic 5: The Supply Constraint — Demand-Led, and It "Will Not Catch Up" in Q4

Intel is running behind an industry-wide shortage across leading-edge wafers, memory, and substrates that management expects to persist "for the foreseeable future." Server is the most constrained line: Zinsner said supply growth is skewed to late Q3 and into Q4, and that even with improvement Intel "will not catch up" this year. The company is pivoting client wafers toward the more-constrained server business to close as much of the gap as possible.

"While things will improve, we will not catch up. We will be behind in the fourth quarter." — David Zinsner, CFO

Assessment: Demand exceeding supply is the best operating condition a capital-intensive manufacturer can have, and it underwrites pricing power and high-confidence capex. It also, however, caps near-term revenue at Intel's own throughput and means the reported growth understates true demand. For the stock, the constraint is a double-edged datapoint: bullish on demand, but a reminder that the upside is execution-gated and that the outsized margins partly reflect a scarcity that will normalize as supply catches up.

Topic 6: Purpose-Built Silicon and the ASIC Push — $2B Run-Rate, $4B in Sight

The design-services and ASIC business is emerging as a genuine third leg. Design-services revenue grew nearly 3x year-over-year, and management put the purpose-built/ASIC business at approximately a $2B run-rate today (up from about $1.2B last quarter), targeting $4B "in the not-too-distant future" against a TAM it sizes above $100B. New wins named this quarter include Fortinet's Security Processor 6 and a multi-year SambaNova collaboration on disaggregated inference, alongside the existing IPU business with hyperscalers.

"Today, we're probably running at about a $2 billion run rate or at least approaching a $2 billion run rate for that business. We think in the not-too-distant future, we'll be at a $4 billion run rate." — David Zinsner, CFO

Assessment: This is a legitimately attractive optionality: it leverages Intel's IP portfolio, advanced packaging, and leading-edge wafers, and it is growing fast off a real base. A doubling to $4B would be material. The margin profile is still undisclosed, and at current scale it is not yet a valuation driver on its own, but it is the most credible new growth vector management surfaced and worth tracking as a swing factor for the 2027-2028 model.

Topic 7: The Government Stake and the Escrowed Shares Drag

The $11.0B GAAP net loss traces almost entirely to a ~$12.5B non-cash mark-to-market on the Escrowed Shares liability owed to the US government under the CHIPS Act Secure Enclave agreement. Because the liability is settled in Intel stock, its value rises with the share price, so a year of enormous share-price appreciation produced an enormous GAAP charge. The balance sheet carries the mirror image: retained earnings fell to $33.8B from $49.0B, and non-controlling interests rose to $15.6B.

Assessment: The charge is non-cash and correctly excluded from the operating picture, but it is not cosmetic. It keeps GAAP results structurally noisy, it drives the "Intel posts $11B loss" headlines that muddy the popular narrative, and it is a standing reminder that the government backstop, which de-risked the balance sheet in 2025, carries an ongoing and counterintuitive cost that scales with success. Model it out of earnings; do not forget it exists.

Topic 8: Memory Ambitions and the Seok-Hee Lee Hire

Management flagged memory as both a supply bottleneck and a strategic area of interest, noting the hire of Seok-Hee Lee, former CEO of SK Hynix, to lead advanced packaging, and gesturing at internal work on compute-memory integration and stacking. Tan framed this as early-stage and collaboration-first with the three major memory vendors, whose supply is a top priority for serving customers.

"Intel has a rich history in the memory. And recently, we hired Seok-Hee Lee to join us. He used to be the CEO of SK Hynix... memory, it becomes the bottleneck, a lot of AI infrastructure and pain point for customers." — Lip-Bu Tan, CEO

Assessment: This is a "stay tuned" item, not a modelable one. A senior memory hire and vague integration commentary are worth noting as directional intent, but there is no revenue, roadmap, or capital commitment attached. It reads as optionality on the far horizon and should carry zero weight in the current valuation.

Guidance & Outlook

The Q3 guide is genuinely strong on the P&L: a $16.3B revenue midpoint sits 7.8% above Street and implies revenue still accelerating, non-GAAP gross margin ticks up to 42.0%, and the $0.38 EPS guide is roughly 41% above the $0.27 consensus. Management framed the sequential shape as more back-half-weighted, with supply improvement skewed to late Q3 and into Q4, and repeatedly noted it will remain behind demand even in Q4. The one guidance line that moved against the equity is capex: raising the 2026 outlook above $20B and flagging 2027 "significantly above" pushes the free-cash-flow crossover further out.

Implied ramp: H1 revenue was roughly $29.7B ($13.6B plus $16.1B); the $16.3B Q3 midpoint puts nine-month revenue near $46B, with a Q4 that management strongly implied steps up further as supply loosens. That points to a full-year figure comfortably above $60B, well ahead of where the Street sat entering the year, and a non-GAAP EPS build of $0.29 plus $0.42 plus a $0.38 guide equals $1.09 through nine months, framing full-year non-GAAP EPS around $1.50.

Street position and guidance style: Consensus was well below both the print and the guide, and the sell-side spent mid-July racing price targets higher (into the $115-155 zone) while mostly keeping Hold and Neutral ratings. That combination, marking targets up to the tape without raising conviction, is itself a signal that the Street sees the stock as fairly valued at best. The guidance style remains one-quarter-forward and, on the P&L, appropriately confident given seven straight beats; the deliberate refusal to guide capex for 2027 beyond "up" is the one place management is holding back a number that clearly exists.

Analyst Q&A Highlights

When Foundry Confidence Becomes Named Customers

The most important line of questioning pressed management on the gap between its stated confidence in external foundry engagement and any actual named commitments, and asked how much of the higher capex is for external customers versus internal needs. Management answered the confidence question at length and the commitment question not at all, describing positive engagement and rising excitement as customers see the 0.9 PDK, without naming a single external customer or quantifying external demand.

Q: "You mentioned increased confidence in engaging with external foundry customers. And I'm curious, when will that confidence be backed by actual customer announcements?"
— Vivek Arya, Bank of America

A: "The engagement with the customer, the feedback have been very positive, tremendous demand for our own products and also external foundry customer engagement that give me the confidence... the moment they're starting to see the 0.9 PDK the yield, they're starting to get excited about what kind of product they want to run that and how much capacity we can provide them. So those are very positive signs that they are really serious about going forward." — Lip-Bu Tan, CEO

Assessment: This is the exchange the thesis turns on, and the answer is the same qualitative reassurance investors heard last quarter, minus the one thing last quarter had: a named anchor. "Really serious about going forward" is not a commitment, and after a quarter in which the Tesla/Terafab relationship went unmentioned, the absence of any external name plus $293M of external revenue is the clearest evidence that the merchant-foundry inflection remains a future event, not a present one.

The Capex Step-Up as a Read on Customer Confidence

A recurring line of questioning probed what the $3B-plus 2026 capex increase, and a "significantly higher" 2027, signals about committed demand, and whether it implies foundry orders. Management framed the raise as an inverse tell: it does not deploy capital ahead of customer confidence, so the raise itself is the signal.

Q: "I'm curious if there's a framework in your mind as you balance meeting customer needs as well as your desired free cash flow goals... if you've got the signed kind of contracts with customers, you're just going to build it?"
— C.J. Muse, Cantor Fitzgerald

A: "We're just going to be very careful around making bets ahead of customer commitments. I think that's the most significant change with Lip-Bu is until we really know that we've got the customers, we don't want to put a significant amount of capital. And you can read that now inversely, given our confidence around next year that we must have pretty significant confidence in our customers." — David Zinsner, CFO

Assessment: The "read it inversely" logic is genuinely reassuring on capital discipline and is the right framework for a company that over-built in the prior era. It does not, however, tell an investor whether the demand behind the capex is internal or external, which is precisely the distinction that determines whether Foundry ever becomes a merchant business. Confidence in demand is not the same as confidence in third-party demand.

Client Strength — Pricing, Not Units

Analysts flagged surprise at client strength given a softening PC end-market and asked whether the revenue upside was pricing or demand. Management was candid that it was largely ASP, from mix and from passing cost inflation through, and guided client roughly flat next quarter with the end-market softer beneath the surface.

Q: "Was it all pricing that drove that revenue strength? Or was there something else going on? And... what your expectations are for client in the second half relative to the end market that sounds like it's going to be worse than typical?"
— Stacy Rasgon, Bernstein Research

A: "I would say it was largely ASP, of which some of that was mix related. Some of that was our own like-for-like changes in ASPs where we thought we had seen some inflation on our cost and needed to pass that on to the end customer. When you look at it year-over-year, it's definitely down that market." — David Zinsner, CFO

Assessment: An honest answer that appropriately discounts the quality of the client beat. Pricing through inflation into a weakening TAM is a margin lever, not a demand signal, and management's plan to divert client wafers to server confirms that client is being managed for cash and capacity rather than growth. It is the lower-quality half of the revenue beat.

The Q4 Step-Up and Undershipping the Market

A pointed line of questioning noted that capacity coming online late in Q3 implies a large Q4 revenue step-up, particularly if Intel is still undershipping true demand by more than $1B, and asked management to confirm the shape. Management leaned into the direction while declining to guide the magnitude, and stressed the constraint will persist.

Q: "That implies a pretty big step-up in revenue for Q4. If I assume that you're still kind of undershipping the market... then it would seem to suggest that Q4 is going to be up pretty big. Is that the right way to think about it?"
— Timothy Arcuri, UBS

A: "If we are able to start to see inventory improve or supply improve towards the end of the third quarter and into the fourth quarter, we would expect a lift from that. I would point out that while things will improve, we will not catch up. We will be behind in the fourth quarter." — David Zinsner, CFO

Assessment: The exchange confirms the demand is real and the near-term ceiling is Intel's own supply, which is bullish for the trajectory and consistent with the raised capex. It also means reported revenue understates demand and that the growth is throughput-gated, so the upside is an execution question, not a demand question. That is a good problem, but it is not the same as unconstrained growth.

Server Share vs. AMD and ARM

Questioners asked how Intel thinks about regaining lost server share over the next five years against both AMD and ARM-based competition, given that it now has internal fabs as a potential advantage. Management acknowledged it is still behind in places and framed the response around the product roadmap and an intent to leapfrog on architecture.

Q: "What do you think happens to your market share in the server space?... as you think about the next 5 years, your competition versus both AMD and also versus ARM, just how are you guys thinking about the prospects for regaining lost share?"
— Joseph Moore, Morgan Stanley

A: "We have a strong road map. We have Clearwater Forest, Diamond Rapid and also Coral Rapid with the SMT involved... Some areas we are still behind, but we are catching up very fast, and we try to leapfrog some of the CPU architecture." — Lip-Bu Tan, CEO

Assessment: The candor ("some areas we are still behind") is characteristic of this management team and is more credible than a claim of leadership would be. Simultaneous multithreading returning at Coral Rapids is the concrete competitive response to per-socket TCO pressure from AMD and ARM. But "catching up" and "leapfrog" are roadmap aspirations for 2027-2028 products; the share-gain thesis is a multi-year bet on execution, not a present-quarter fact.

ASIC Growth and Margin Profile

A question sought the diversity, growth profile, and margin profile of the purpose-built/ASIC business following the Fortinet win. Management sized the run-rate and target enthusiastically while sidestepping the margin question entirely.

Q: "Based on what was disclosed last quarter, it's about a $1.2 billion run rate business now growing well for the company. How do we think about the diversity of that business?... And maybe also with that, the margin profile of ASICs as they expand."
— Aaron Rakers, Wells Fargo

A: "We're probably running at about a $2 billion run rate or at least approaching a $2 billion run rate for that business. We think in the not-too-distant future, we'll be at a $4 billion run rate... we think we deserve a good chunk of that given our IP portfolio and what we bring to bear. So stay tuned." — David Zinsner, CFO

Assessment: A near-doubling of the run-rate in two quarters is a real acceleration and the most tangible new growth vector on the call. The unanswered margin question matters, though: ASIC/foundry-style economics are typically lower-margin than Intel's product business, so a fast-growing ASIC mix is a revenue story whose profit contribution is unproven. Track the run-rate; withhold the margin credit until disclosed.

Funding the Build-Out — The Balance Sheet

A follow-up asked whether product-side cash generation is enough to fund the stepped-up back-half and 2027 investments, or whether something else is required. Management pointed to $40B of liquidity and roughly $10B of monetizable non-core assets, then acknowledged the possibility of tapping the capital markets if the growth materializes.

Q: "As you are planning these investments for the back half of the year and into next year, how are you thinking about the balance sheet? Do you think that the success you are seeing in your product business... is enough to kind of fund a lot of these investments? Or will something else be required?"
— Vivek Arya, Bank of America

A: "We have over $30 billion of cash. We have a $10 billion revolver... roughly, call it, $10 billion of what I call noncore assets that can still be monetized... That said, if we're super successful, which we're driving to, we may need to tap the capital markets to drive some more investment." — David Zinsner, CFO

Assessment: The liquidity position is strong and the balance sheet is no longer the risk it was in 2024. But the explicit acknowledgment that success could require additional capital-market financing is a meaningful caveat for a 90x-forward equity: it flags potential future dilution or leverage precisely when the capex line is rising and free cash flow is negative. This is the cash-flow counterweight to the P&L story, stated by the CFO himself.

What They're NOT Saying

1. The Tesla/Terafab 14A anchor: Last quarter's single most important disclosure, the named 14A anchor customer that underpinned our maximum-conviction call, was not mentioned once. Tesla, Terafab, SpaceX, and Musk were all absent from a call that discussed 14A extensively. Silence is not cancellation, but the disappearance of a marquee relationship one quarter after its introduction is the most conspicuous omission in the release.
2. Any named external foundry customer: Beyond the ASIC wins (Fortinet, SambaNova), no external wafer foundry customer was named, and external foundry revenue of $293M was disclosed only in passing. The merchant-foundry demand case remains asserted rather than evidenced.
3. The 2027 capex number: Management confirmed 2027 capex will be "significantly above" 2026's $20B-plus but declined to quantify it, deferring to "the very beginning of the year." A number this material being withheld leaves the free-cash-flow model unanchored on its single largest swing factor.
4. Foundry breakeven year: Still not recommitted. The Q3 2025 "as we exit 2026" framing has not been repeated for two consecutive calls. A formal breakeven target remains the disclosure that would most help the merchant-foundry valuation case, and it continues to be absent.
5. ASIC margins: Management sized the ASIC run-rate ($2B toward $4B) but explicitly declined to characterize its margin profile, leaving open whether the fastest-growing new vector is accretive or dilutive to corporate margins.
6. No full-year revenue guide: Despite seven straight beats and clear multi-year visibility from long-term agreements, management held the one-quarter-forward line. Consistent with prior discipline, but it forces investors to build the second half from segment commentary.
7. The path to positive free cash flow: Zinsner allowed that reaching positive free cash flow next year is "a little bit more challenged" given back-end investment, without offering a timeline. For a stock priced on future cash generation, the deferral of the FCF crossover is a quiet but important omission.

Market Reaction

• Pre-print setup: INTC closed the regular 2026-07-23 session at $100.23, down about 2.3% intraday ahead of the print. It entered the quarter up roughly 172% year-to-date and 327% over the trailing twelve months, having peaked near $141 and then corrected roughly 28% into the print (down about 24% in the trailing thirty days alone). It traded near 90x forward non-GAAP earnings, with prediction markets reportedly pricing near-certain odds of a beat.
• After-hours move: Shares initially jumped as much as +12-13% within minutes of the release crossing the wire, then gave back more than half of that pop before the next open as the call and the $11B GAAP-loss headlines circulated.
• Reaction session (2026-07-24): INTC closed at $97.19, down 3.0% ($3.04), with an intraday range of $95.37 to $101.74. Volume of 88.6M was 0.7x the 30-day average, so this was a fade on unremarkable volume, not a capitulation.
• Benchmark: The S&P 500 rose about 0.6% on the reaction day, so INTC underperformed the tape by roughly 360 bps on the session.

The reaction is the analysis. A two-times EPS beat, an 8%-above-Street guide, and a de-risked roadmap that produces a 3% decline is a positioning-and-valuation event, not a fundamentals event. The contrast with Q1 2026 is instructive: a comparably outsized surprise in April drove a roughly +23.6% intraday move, because the stock was cheap and under-owned coming out of the Q4 guide-miss sell-off. Three months and a near-tripling later, the same magnitude of beat cannot move the shares, because everything the beat contained is already capitalized. When good news stops working, the risk/reward has changed even if the business has not.

Street Perspective

Debate: Is a Fully-Priced Turnaround Still a Buy?

Bull view: Forward earnings power is inflecting faster than the multiple implies; at a full-year non-GAAP EPS trajectory near $1.50 and a plausible $2.25-2.75 in 2027, the stock is closer to the high-30s on next-year earnings than the headline 90x on trailing. A growth-reaccelerating leader with US industrial-policy support and foundry optionality deserves a premium multiple.

Bear view: The stock is up 327% in a year and faded its best quarter; the sell-side raised targets to $115-155 while keeping Hold and Neutral ratings, which is marking-to-market, not conviction. At 90x forward with negative free cash flow, a Foundry losing $2.1B a quarter, and rising capex, the reward for the turnaround being right has already been paid.

Our take: The bear framing is the stronger one for the stock, even as the bull framing is the stronger one for the business. Forward earnings are indeed rising fast, but a company still generating GAAP losses and negative free cash flow, at a full multiple even on optimistic 2027 numbers, offers balanced-at-best twelve-month risk/reward. This is the "strong beat, full valuation, Hold" case in its purest form.

Debate: Is the Foundry Merchant-Business Inflection Real or Perpetually Deferred?

Bull view: Committing 14A to 2028 HVM, 18A output 25% above target, PDK 0.9 on track for October, and defect density ahead of 18A are all evidence the technology is landing; external customers will follow the proof, and the ASIC business is already a $2B run-rate leg.

Bear view: External foundry revenue is $293M, no external wafer customer was named, the prior quarter's marquee anchor went unmentioned, and the breakeven year has not been recommitted for two calls. The merchant-foundry story has been "next year" for several years.

Our take: The node is increasingly credible; the merchant business is not yet a business. Committing to 2028 HVM de-risks the roadmap, but the equity is being asked to pay today for external demand that is still, on the company's own disclosure, an engagement rather than a book. We treat Foundry as validated on cost and technology, and unproven on third-party revenue. That distinction is exactly why conviction should come down, not up.

Debate: How Durable Is the DCAI Margin?

Bull view: DCAI at a 40% operating margin on 59% growth reflects genuine operating leverage, rising core counts, and durable AI-driven CPU attach; as Intel 3 and 18A server capacity scales, the segment compounds from here.

Bear view: Part of that 40% margin is scarcity rent on a supply-constrained peak; as supply catches up in 2027 and as AMD and ARM press on per-socket TCO, both the growth rate and the margin normalize toward more typical server-CPU economics.

Our take: Both are partly right. The demand reacceleration is structural and the CPU-attach thesis is confirmed, but a 40% segment margin struck while the company cannot fully supply the market is unlikely to be the steady-state. We would model DCAI growth decelerating from the current pace and margins easing modestly as supply normalizes, which still leaves a far healthier segment than the 2024-2025 bear case assumed, just not one that compounds at Q2's rate indefinitely.

Model Implications

Valuation: The earnings estimates move up meaningfully, and yet the rating moves down, which is the entire point. On our revised framework, FY26 non-GAAP EPS lands near $1.50 and FY27 near $2.25-2.75. At the reaction-day close of $97.19, that is roughly 65x FY26 and about 35-43x FY27 non-GAAP earnings. Even crediting the fast forward growth, mid-to-high-30s on next-year earnings is a full multiple for a company still posting GAAP losses, burning free cash flow, and carrying a Foundry that loses $2B a quarter. A fair-value band of roughly $85-110 captures the reasonable range: the low end applies ~32x to FY27 EPS with the FCF and Foundry overhangs weighted; the high end applies ~40x to the upper EPS case if DCAI momentum and ASIC scaling continue. At $97, the stock sits in the middle of its own fair-value range, which is the definition of a Hold. The move from Outperform is not a downgrade of Intel; it is the recognition that the market has finished paying us for the thesis being right.

Thesis Scorecard Post-Earnings

Overall: Thesis operationally strengthened, investment case weakened. Every fundamental pillar that concerns the business is intact or better: 18A is ramping, DCAI is inflecting, margins are in the 40s, and 14A is committed to 2028 HVM. But the two pillars that concern the stock, valuation and free cash flow, have both moved the wrong way, and the one pillar that most needed to advance (external foundry demand) went backward with the disappearance of the Terafab anchor. A proven turnaround at 90x forward, fading its best quarter, is a different risk/reward than an unproven turnaround at 10x was.

Action: Trim into strength; do not add. Positions established in the $19-40 range from Q3 2025 through Q4 2025 carry enormous embedded gains and can be scaled back to a core holding here. Fresh capital should wait: a pullback toward the low $80s, or hard evidence that Foundry's external revenue and the free-cash-flow trajectory are inflecting rather than just the P&L, would restore Outperform. Until then, the business is winning and the stock has already won.