I'll research Intel before writing the preview. Let me gather recent earnings materials, news, and price data.Let me read the most recent earnings materials and pull recent news.The data is on a forward-shifted timeline (we're at July 2026; last report was Q1 2026). Let me read the latest earnings call transcript and search for recent news.Let me check for recent 8-Ks and press releases since the last report for any material developments.# Intel (INTC) — Q2 2026 Earnings Preview

Report date: Thursday, July 23, 2026, after market close · Call: 2:00 p.m. PDT / 5:00 p.m. ET · Webcast at intc.com

---

The one-line setup

Intel enters this print with more momentum than it has had in years. Management has now beaten guidance for six straight quarters, the narrative has flipped from "can Intel survive?" to "how fast can Intel add capacity?", and demand is running ahead of supply across every business. The debate for investors is no longer solvency — it's the durability of the CPU-in-AI thesis, the gross-margin trajectory as 18A ramps, and whether the supply constraints that are currently capping revenue can be unlocked without blowing up costs.

What's guided for Q2 2026

Management's own Q2 outlook (given April 23) is the baseline to beat:

Metric (Q2'26 guide)	GAAP	Non-GAAP
Revenue	\$13.8B – \$14.8B (midpoint \$14.3B, +2–9% QoQ)	—
Gross margin	~37.5%	~39.0%
Tax rate	~4%	~11%
EPS (diluted)	\$0.08	\$0.20

Key framing from the CFO: DCAI is expected up double digits sequentially, CCG also up on better supply plus a full quarter of pricing actions. Gross margin steps down modestly from Q1's 41% because of (1) a much larger Intel 18A mix contribution (still early, dilutive) and (2) Q1 inventory benefits that don't repeat.

Given the six-quarter beat streak and Q1 coming in $1.4B above the midpoint with non-GAAP EPS of $0.29 vs. a ~breakeven guide, expectations for another beat are effectively baked in — the risk is that the bar has quietly risen.

Where the stock is coming in

• INTC is up roughly 16% year-to-date (~$20.22 → ~$23.51), and about +30% off its April low (~$18) — but still ~14% below its February peak (~$27+).
• That YTD gain badly lags the AI-levered peers: AMD +~32%, NVDA +~24% over the same stretch. The market is rewarding the turnaround but is not yet paying an "AI winner" multiple. That gap is the crux of the bull case and the reason a soft print could be punished.

The five things that actually matter on this call

1. Data Center (DCAI) and the "CPU is back" thesis. This is the whole story. In Q1, DCAI grew +22% YoY / +7% QoQ to $5.1B with operating margin of 31%. Management's structural argument: as AI shifts from training → inference → agentic, the CPU-to-GPU ratio is moving back in the CPU's favor (from ~1:8 toward 1:4 and potentially parity). "AI-driven" businesses are now ~60% of revenue and grew ~40% YoY. Watch for: another double-digit sequential DCAI quarter, updated commentary on the server unit-TAM outlook (guided to double-digit unit growth for 2026, extending into 2027), and any new long-term agreements (LTAs) beyond the Google Xeon + custom IPU deal.

2. Supply constraints — how big is the "gap"? Management said it is undershipping demand by an amount that "starts with a B" (billions). Every incremental point of output flows straight to revenue and margin. Listen for: are yields/cycle-time gains and wafer-start increases closing the gap faster or slower than the "measured pace" flagged for 2H? This is the single biggest swing factor on both the top line and the beat.

3. 18A / Panther Lake ramp and the margin path. Intel 18A yields are reportedly running ahead of internal targets (a target for year-end being hit by mid-year), but 18A is still a gross-margin headwind early in the ramp. Panther Lake (Core Ultra Series 3) volume was expected up 6–7x sequentially in Q2, at below-corporate-average margins. Key question: does the full-stack (product + foundry) margin on 18A improve enough to validate the path back to Intel's target model?

4. Foundry and 14A external traction. Intel Foundry posted $5.4B revenue (+16% YoY) but external revenue was tiny ($174M) with a $2.4B operating loss. Management is deliberately stepping up 14A investment for customer evaluations and expects design commitments to emerge in 2H26 into 1H27. Advanced packaging backlog is described as "billions per year." Any concrete external 14A customer win — or new color on the Terafab partnership with SpaceX/xAI/Tesla — would be a meaningful catalyst.

5. PC (CCG) and the 2H air pocket. CCG was $7.7B (+1% YoY) with a strong 33% operating margin. But management is prudently planning for PC demand to weaken in 2H, with the full-year PC unit TAM down low-double-digits. Intel expects to be less impacted than the TAM (pricing + channel inventory replenishment), modeling client revenue roughly flat from Q2 onward. Watch for any change to that cautious framing.

Cost & margin cross-currents to listen for

• Rising input costs — memory, substrates, and T-glass — are a growing 2H gross-margin headwind that partially offsets yield gains. Memory in particular is a wildcard given industry-wide price inflation.
• OpEx was directionally targeted at ~$16B for 2026 but management flagged it will likely run higher (inflation, variable comp, targeted investments).
• CapEx for 2026 was nudged to flat YoY (from flat-to-down), with tool spend up ~25% (growth-oriented) even as fab "shell" spend comes down.

Balance sheet & capital allocation

• Cash + short-term investments ~$32.8B; total debt ~$45B.
• Intel bought back the 49% minority stake in the Fab 34 (Ireland) JV, funded with ~$7.7B cash + $6.5B of new senior notes (issued across five tranches in April). This is accretive but consumed liquidity and adds leverage — expect questions on the resulting NCI (~$250M/quarter Q2–Q4) and free-cash-flow trajectory (management still guides to positive adjusted FCF ex-Fab 34).
• The U.S. government equity interest (CHIPS Act "Secure Enclave" / escrowed-shares structure) remains a mark-to-market swing item in GAAP results and a governance overhang worth monitoring.

Governance / housekeeping flags

• Chief Accounting Officer Scott Gawel resigned on April 24 (the day after Q1); CFO Dave Zinsner assumed the principal-accounting-officer role on top of CFO. A dual-hatted CFO/CAO is worth keeping an eye on, though framed as a departure "to pursue another opportunity."
• At the May 15 annual meeting, shareholders backed management on pay/equity plans and rejected activist proposals — reducing near-term activism risk.

Bottom line — how to read the print

• Bull case validation: Revenue above the ~$14.3B midpoint, DCAI up double digits again (with new LTAs), non-GAAP GM holding ≥39% despite the 18A mix, and evidence the supply gap is closing — plus any external 14A/foundry design win.
• Watch for disappointment if: the beat is smaller than the six-quarter cadence has trained investors to expect; 2H gross-margin guidance is cut on memory/substrate cost inflation and 18A dilution; PC weakness looks worse than "less impacted than TAM"; or foundry losses widen with no external customer proof points.

Net: the fundamentals are inflecting and the "CPU in the AI era" story has real data behind it, but with a beat essentially priced in and margins pressured by the very ramp that's driving growth, the reaction will hinge on the 2H margin guide and tangible foundry/14A external progress more than on the Q2 headline itself.

Note on data: figures above are drawn from Intel's Q1 2026 earnings release and call, related 8-K filings, and recent price action; all forward metrics are management guidance, not consensus, which I couldn't independently pull from the tools available.