INTC — Q2 2026 Earnings Preview
Setup in one line
Four semis/foundry peers report ahead of Intel — ASML (7/15), TSMC (7/16), TXN (7/22) and STM (7/23, hours before) — setting the AI/datacenter and foundry-utilization tone. Management walks in on offense after a Q1 print the stock rewarded sharply, having framed the story as shifting from “can we survive” to “how fast can we scale supply.” But the aggressive tone is now fully priced and then some: a July selloff on reports that 18A profitable yields slip to late-2026/2027 and on AMD reportedly passing Intel in quarterly data-center revenue means 7/23 is a referendum on concrete 18A/14A/DCAI specifics — with consensus already at the top of the guide, an in-line print with unchanged foundry timelines could still disappoint.
Intel is a turnaround-in-progress, not yet a leader-stays-leader story: it has ceded the AI-accelerator crown to NVIDIA and, per July reporting, just lost the quarterly data-center revenue lead to AMD — so this print is judged on whether the CPU/foundry recovery under CEO Lip-Bu Tan (since Mar 2025) is durable, not on category dominance. The P&L is a function of (a) CCG/client volume against a weakening PC TAM, (b) DCAI/server-CPU reacceleration, and (c) Intel Foundry's ramp economics (18A mix today, 14A external customers as the make-or-break future).
Growth trajectory — reaccelerating, supply-constrained. The last print (Q1 2026, reported 2026-04-23) beat across the board: revenue $13.58B (+9.3% vs the guide midpoint, $1.4B above its own mid), non-GAAP EPS $0.29 vs a breakeven guide, and DCAI $5.05B (+22% YoY) as the standout. Management called it the sixth consecutive quarter of exceeding financial expectations and reframed the constraint as supply, not demand — unmet demand "starts with a B" (>$1B).
Key watch items into Q2 2026:
- Guidance / the bar: the Q2 guide is revenue $13.8B–$14.8B (mid $14.3B), non-GAAP GM 39%, EPS $0.20, tax 11%. Consensus ($14.42B rev / $0.21 EPS) sits above the midpoint — the tell of a chronic sandbagger, but also a bar with little cushion for an in-line-to-soft read.
- Gross margin (the contested line): GM guided down to 39% from Q1's 41% actual on a "meaningfully larger" 18A mix (Panther Lake volume up ~6–7x QoQ, below-corporate margin) plus non-repeating Q1 inventory benefits. Memory/substrate/T-glass cost inflation is the self-flagged H2 headwind now materializing (DRAM contract prices +13–18% QoQ into Q3).
- DCAI vs AMD: guided "up double digits" sequentially — the single most-watched line, directly against reports that AMD EPYC passed Intel in data-center revenue.
- 18A yields & 14A external customers: the July stock scare was about 18A yield economics (profitability timing), not feasibility; KeyBanc pegs 18A yield ~85%. 14A external commitments remain a 2H26→1H27 story (still ~zero committed) — a progress update this quarter, not a resolution. Watch for an Investor Day date (promised H2 2026, still unannounced).
Classification: CONSERVATIVE guider, consistent beater — but priced for perfection on tone. Six straight above-high-end quarters make a Q2 beat the base case; the risk into 7/23 is that the revenue-surprise magnitude compresses (Street already near the guide high end) and that GM, not revenue, becomes the swing factor.
The Q2 2026 guide was set on the Q1 call (2026-04-23) by CFO David Zinsner: "we're guiding Q2 revenue to a range of $13.8 billion to $14.8 billion, up 2% to 9% sequentially… At the midpoint of $14.3 billion, we forecast a gross margin of 39%, a tax rate of 11% and EPS of $0.20, all on a non-GAAP basis." Intel guides the current quarter only — no numeric Q3 or full-year revenue/EPS range.
| Q2 2026 metric (non-GAAP) | Guide (mid) | Guide range | Consensus | Read-through |
|---|---|---|---|---|
| Revenue | $14.3B | $13.8B–$14.8B | $14.42B | Street +0.8% above mid (upper half of range) — a small beat is already baked in |
| Non-GAAP gross margin | 39.0% | not ranged | ~39% | Down from Q1's 41% on 18A mix + no repeat inventory benefit — the contested line |
| Non-GAAP EPS | $0.20 | — | $0.21 | Street +$0.01 above guide; clean-EPS beats have trended +21c→+27c |
| Tax rate | 11% | — | n/a | Guided; low-variance |
Segment color embedded in the guide (directional, not numeric): sequential revenue growth in both CCG and DCAI, with DCAI up double digits, on improving available supply plus a full quarter of pricing actions. Full-year framing: PC unit TAM down low-double-digit % YoY; client revenue "flattish from Q2 onward"; H2 to follow the 10-year seasonal pattern (servers up, PCs down). No Q3 numeric guide is given.
3a. Current quarter (Q2 2026) — guide vs. year-ago comp
Only consolidated revenue is numerically guided; segments are guided directionally (marked accordingly). Revenue ≈ units × ASP × mix. The Q2 2025 base was depressed by restructuring/impairment charges (GM 29.7%, EPS −$0.10), so the YoY comparison flatters — the cleaner signal is the +11.2% revenue-midpoint growth, the first double-digit-guided quarter of the Lip-Bu Tan era.
| Metric | Q2'25 actual (comp) | Q2'26 guide / cons. | Implied YoY (mid) | Framing |
|---|---|---|---|---|
| Revenue | $12.86B | $14.3B mid / $14.42B | +11.2% | Range +7.3% (low) to +15.1% (high) YoY |
| Non-GAAP gross margin | 29.7% (charge-hit) | ~39% | +~930bps | Down seq from Q1 41% on 18A mix; the swing line |
| Non-GAAP EPS | −$0.10 | $0.20 / $0.21 | +$0.30 swing | Loss → profit; year-ago hit by charges |
| CCG revenue | $7.87B | up seq (>~$7.73B); dir. | ~flat-to-up | Full qtr of pricing; Panther Lake ramp dilutive to GM |
| DCAI revenue | $3.94B | up double-digit seq (≥~$5.56B); dir. | ~+40%+ | The most-watched line; vs AMD share-gain reports |
| Intel Foundry revenue | $4.42B | not guided | — | External only ~$174M in Q1; op loss −$2.4B |
3b. Historical quarterly trend (Daloopa) — trajectory over absolutes
Interpretation: revenue has stepped up four straight quarters and DCAI is the acceleration engine ($5.05B, +22% YoY in Q1'26). The Q1'26 non-GAAP EPS of $0.29 is the biggest clean beat of the recovery. *Q2'25's −$0.10 was charge-driven; ex-charge it was ~$0.10 and above guide. The tell into Q2'26 is whether DCAI's guided double-digit sequential growth lands and whether GM holds the 39% bar — not the revenue headline.
3c. FQ+1 (Q3 2026) and FY+1 (FY2027) — no company guide
Intel does not issue numeric guidance beyond the current quarter. Q3 columns are street/derived; FY rows are consensus.
| Period | Revenue (cons.) | Non-GAAP EPS (cons.) | # analysts | Note |
|---|---|---|---|---|
| Q3 2026 | ~$15.4B (derived) | n/a (not isolated) | — | Derived from FY26 street minus H1; H2 seasonal (servers up, PCs down). Memory/substrate GM headwinds flagged |
| FY2026 | $58.7B | $1.088 | 24 | Company gave no FY revenue/EPS guide; OpEx ~$16B "likely higher", CapEx flat YoY |
| FY2027 | $65.7B | $1.596 | 24 | Consensus-only; +12% rev / +47% EPS on foundry-loss narrowing + server mix |
The setup in one paragraph: management enters Q2 having escalated its tone sharply into offense over the last three calls, and the aggressive bar is now fully priced — and then some. The Q2 guide rests on DCAI double-digit sequential growth, a full quarter of pricing, and rising supply, but three self-flagged headwinds are materializing exactly as warned: 18A mix (dilutive to GM), memory/substrate cost inflation, and a PC TAM planned down low-double-digits in H2. A July selloff reframed the Q1 "18A yields running ahead" message as a profitability delay to late-2026/2027, and reports that AMD passed Intel in data-center revenue hit the DCAI-strength thesis directly. The core recovery is intact, but 7/23 is a referendum on specifics, not confident tone.
Tone trajectory (one-way escalation): Q3'25 survival / balance-sheet repair ("still a long way to go") → Q4'25 cautious progress ("humble… never satisfied") → Q1'26 offense / scaling ("A year ago the conversation was whether we could survive; today it's how quickly we can scale supply"). Guidance-language inversion is the confidence tell: Q4'24 management refused to guide beyond one quarter; by Q1'26 they volunteered full-year framing. The contrarian angle (per principles): management loudly asserts the CPU is re-centering in the AI stack (CPU:GPU ratio moving 1:8 → 1:4 → toward parity) while the Street still frames Intel as an accelerator also-ran — a management-vs-street disconnect worth testing on the call.
Management-quality read: guidance accuracy has been excellent to the upside — 6/6 revenue-midpoint beats, EPS routinely multiples of a breakeven guide. This is a team that under-promises and over-delivers, and whose forward claims (18A yields "ahead of internal projections," DCAI momentum "into 2027") have so far been corroborated by results. The honest-broker caveats they keep flagging — foundry profitability is a multi-year journey, memory/substrate inflation "could impact demand," PC down in H2 — are the items to hold them to.
| Guidance driver | Confidence into Q2 | Basis / caveat |
|---|---|---|
| DCAI double-digit seq growth | High | Multiple 3–5yr LTAs (Google named); Xeon 6 host CPU for NVIDIA DGX Rubin NVL8; ASIC run-rate >$1B. But AMD reportedly passed Intel in DC revenue |
| Server CPU demand into 2027 | High / rising | Outlook "improved over the last 90 days"; "double-digit unit growth… into 2027" |
| Revenue midpoint | Medium-high | Governed by supply execution (yields, cycle time, wafer starts) — in management's control but still ramping |
| Gross margin ~39% | Medium | Pricing firm; 18A mix the swing factor. Zinsner "roughly in the ZIP code of Q1" |
| PC / CCG revenue | Medium-low | Consumption TAM down low-double-digits H2; cushioned by pricing + inventory replenishment, but flagged demand-risk if component inflation bites |
| 18A yield / foundry margin path | Low-medium (key debate) | Q1 said yields "ahead of internal projections"; foundry-average GM "multiple, multiple quarters" away; Foundry lost $2.4B in Q1 |
Post-guidance updates since 2026-04-23 (no formal guide revision):
(1) 18A-P entered risk production (VLSI Symposium, ~6/16) — on-timeline execution proof. (2) Computex (6/1): Xeon 6+ on 18A, Crescent Island AI GPU preview, SambaNova/Foxconn rackscale intent — extends the DCAI/heterogeneous-compute narrative. (3) Fab 34 (Ireland) 49% buyout closed (~$7.7B cash + $6.5B debt) plus ~€5B Irish expansion — accretive but adds to debt/NCI (~$250M/qtr Q2–Q4). (4) Investor Day promised H2 2026 at Santa Clara — date still unannounced; a potential catalyst management may set on the call. (5) Memory (DRAM/NAND) prices ~doubled in 2026, DRAM contract +13–18% QoQ into Q3 — the exact H2 GM headwind management pre-flagged, now materializing.
| Catalyst | Q1'26 KPI baseline | What consensus expects in Q2'26 | Direction |
|---|---|---|---|
| DCAI / server CPU reacceleration | DCAI $5.05B (+22% YoY) | Guided up double digits seq — the single most-watched line. Wells Fargo expects a "strong server CPU update" | Watch (AMD) |
| Gross-margin trajectory | Non-GAAP GM 41% in Q1; GM$ $5.35B | Steps down to ~39% (18A mix + no repeat inventory benefit); memory/substrate/T-glass costs build into 2H | Negative |
| 18A ramp / Panther Lake (Core Ultra 3) | CCG $7.73B (+1.3% YoY) | CCG up modestly seq; Panther Lake volume up 6–7x QoQ — a near-term GM headwind (PL margins below corporate avg) | Mixed |
| Intel Foundry 14A external customers | Foundry $5.42B; external only $174M; op loss −$2.4B | No committed external 14A customer expected this quarter; watch PDK progress + Foundry op-loss narrowing. Decisions 2H26→1H27 | Watch |
| Client PC cycle (AI PC vs weak TAM) | AI PC >60% of client CPU mix | FY PC unit TAM guided down low-double-digits, weaker in 2H; Intel less exposed than TAM on pricing + inventory replenishment | Negative |
| Investor Day date + strategic capital | US govt 9.9% stake ($8.9B CHIPS); SoftBank $2B; NVIDIA collab | No new capital event expected; watch for an Investor Day date (promised H2 2026, unannounced) — potential upside catalyst | Watch (upside) |
Bull case
DCAI clears double-digit sequential growth and holds share vs AMD, GM prints at/above 39%, 18A yield/foundry-loss narrative de-risks, and an early 14A commitment signal or Investor Day date lands — revenue beats the $14.42B bar cleanly and the recovery reads durable.
Bear case
In-line revenue (Street already at guide high end), GM slips below 39% on 18A mix + memory-cost inflation, DCAI decelerates against AMD share gains, and 14A stays at ~zero committed with unchanged foundry timelines — a technically fine print still disappoints a ~15% implied move.
Ex-earnings newsflow since the Q1 report (2026-04-23), most material first. The quarter's biggest driver has been unconfirmed foundry design-win rumors (Google TPU, NVIDIA 18A) that fueled a June rally, then a July yield-economics scare — both flagged for verification risk.
| Date | Item | Earnings read-through |
|---|---|---|
| ~Jul 15, 2026 | KeyBanc: Intel 18A yield up to ~85% (from ~65% prior qtr); ASML validates High-NA EUV on Intel line | Directly rebuts the July yield-scare. ~85% is a commercially viable band; if confirmed on the call, de-risks the 18A gross-margin drag thesis. Sell-side estimate, not company-disclosed. |
| ~Jul 8, 2026 | Sharp selloff on reports 18A/18A-P won't hit profitable yields until late-2026/2027; AMD reportedly passes Intel in quarterly DC revenue | Most material negative in the window. Yield-economics, not feasibility, is the debate; reframes Q1's "running ahead" as a delay. Sets a low-expectations bar into the print and pressures the DCAI thesis. |
| Jun 2026 | 18A-P enters risk production (VLSI Symposium); +9% perf iso-power / −18% power iso-perf vs 18A | Company-confirmed, on-timeline execution proof. Credibility-builder for the foundry roadmap ahead of the print. |
| ~Jun 8, 2026 | Google reportedly orders 3M+ TPUs (2028) from Intel Foundry; NVIDIA evaluating 18A + advanced packaging (both unconfirmed) | Biggest catalyst of the quarter — drove the June rally. NOT confirmed by either company; JPMorgan called the TPU story "a storm in a teacup." Watch for any named-customer / committed-volume language vs the ~$174M/qtr external-foundry base. |
| ~Jun 2026 | Computex lineup: Panther Lake shipping in 200+ laptop designs; 18A positioned as foundry calling card | Reinforces 18A as both product and foundry story; supports the CCG ramp but is the near-term GM headwind. |
| Ongoing | Restructuring under Lip-Bu Tan continues — headcount ~125k → <100k; foundry-first strategy | Cost-out program underpins the non-GAAP EPS ramp. Most cuts landed in 2025, so it frames rather than moves the Q2 opex/margin line. |
Reliability read: the highest-signal, hardest-to-verify items (Google-TPU, NVIDIA 18A) drove the round-trip and are the reason management will be pressed on external-foundry commitments on 7/23. Company-confirmed items (18A-P risk production, Panther Lake shipping) are solid. The KeyBanc ~85% yield and the July yield-economics timing debate sit in direct tension — that tension is the Q2 gross-margin question.
Intel is a consistent beater: over the last 12 quarters (Q2'23→Q1'26) revenue beat consensus 10/12 (83%) and non-GAAP EPS beat 9/12 (75%) headline — but 11/12 (92%) ex-charge. Over the last 4 quarters: revenue 4/4 (100%), EPS 3/4 (75%). Both trailing EPS "misses" (and 2 of the 3 in the full 12) were driven entirely by restructuring/impairment charges not in the guide — ex-charge, those quarters beat. Beat magnitude is improving: the last two clean beats (Q3'25 +21c, Q1'26 +27c on EPS; +3.5% and +9.3% on revenue) are the largest of the recovery.
| Metric | Q2'23 | Q3'23 | Q4'23 | Q1'24 | Q2'24 | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Revenue surprise % |
+18.1% | +4.9% | +1.7% | −0.6% | −0.7% | +2.0% | +3.1% | +3.0% | +7.3% | +3.5% | +1.8% | +9.3% |
| Non-GAAP EPS beat, ¢ |
+11¢ | +20¢ | +9¢ | +4¢ | −8¢* | −44¢* | +1¢ | +12¢ | −11¢* | +21¢ | +7¢ | +27¢ |
| Quarter | Rev actual | Rev est. | Rev surprise | EPS beat (¢) | Result |
|---|---|---|---|---|---|
| 2025 Q1 | $12.67B | $12.30B | +3.0% | +12¢ | Beat |
| 2025 Q2 | $12.86B | $11.98B | +7.3% | −11¢* | Rev beat / EPS charge-miss |
| 2025 Q3 | $13.65B | $13.20B | +3.5% | +21¢ | Beat |
| 2025 Q4 | $13.67B | $13.43B | +1.8% | +7¢ | Beat |
| 2026 Q1 | $13.58B | $12.42B | +9.3% | +27¢ | Beat |
| 2026 Q2 (est.) | TBD | $14.42B | — | street $0.21 | Report 7/23 |
Pattern verdict — consistent beater, but the easy revenue beat may compress. Revenue beat the midpoint in all 6 measurable quarters (avg +5.7%); applying that to the $14.3B mid implies a landing near ~$15.1B — above the high end. But the Street is already at $14.42B (≈ guide high end) / $0.21, so a repeat of the +9% surprise requires supply again outrunning plan; the magnitude of the revenue surprise should compress vs Q1'26. Charge risk is lower (restructuring largely complete) but not zero — the swing factor for a clean vs noisy EPS beat is the larger 18A mix + rising memory/substrate costs against the 39% GM bar.