The Numbers
| Metric | Estimate | Actual | Result |
|---|---|---|---|
| EPS | $0.22 | $0.42 | +92% 🔥🔥 |
| Revenue | $13.1B | $14.2B | +8.4% ✅ |
| Foundry | $1.8B | $2.6B | +44% 🔥 |
| Client (PC) | $7.2B | $7.6B | +5.6% |
Market Reaction: -2.46% to $100.10 — sell-the-news despite monster beat. Volume 87M (vs avg 133M).
Guidance Update
Q3 2026
- Revenue: $14.8B–$15.4B (street was $14.2B) ✅
- EPS: $0.55–$0.65 (street was $0.48) ✅
- Foundry gross margin: positive (first time in 3 years)
FY 2026 Raised
- Revenue: $57B–$59B (was $54B–$56B)
- EPS: $2.10–$2.40 (was $1.80–$2.10)
- Foundry profitability by Q4
What Happened
✅ Foundry Business Explodes (+44% Revenue)
- 18A node ramping ahead of schedule
- Microsoft, Amazon, Google orders doubled QoQ
- First-ever positive gross margin projected for Q3
Why It Matters: Foundry was the black hole burning $3B/year. Now it’s a growth driver. If this sustains, INTC’s valuation re-rates.
✅ AI Chips Triple Revenue
- Gaudi 3 accelerator orders: $780M (vs $240M Q1)
- Beating AMD in price/performance for inference
- NVDA still dominates training, but Intel winning inference share
Key Win: Meta ordered 50,000 Gaudi 3 units — largest single deal in Intel AI history.
✅ PC Business Stabilizing
- Client revenue $7.6B (+5.6% YoY)
- Average selling price (ASP) up 8% — premium mix
- Copilot+ PC refresh cycle starting (AI PCs)
Context: PC market was dead 18 months ago. AI PC refresh = unexpected tailwind.
⚠️ Data Center Still Weak
- Server CPU revenue flat YoY
- AMD taking share in hyperscaler (EPYC 4th gen ramp)
- Mitigation: Custom silicon (Foundry) offsets
Why the Stock Dropped
Simple: INTC is up 353% YTD (from $18.97 low to $142.35 high in May). After that run, a 92% beat is priced in.
Profit-Taking Math:
- YTD gain: +253% (even after today’s -2.5%)
- P/E: undefined (barely profitable)
- Valuation stretched vs historical avg (P/S ~3.5x vs 2.0x norm)
Sell-the-news is rational. Stock needed to consolidate.
Why This Matters
1. Foundry Turnaround Is Real
For 3 years, Intel’s foundry burned cash. Now it’s gross margin positive and winning hyperscaler designs. This validates the entire thesis.
2. AI Chip Competition Heats Up
NVDA dominates training (~90% share). But inference is the next battleground — and Intel’s Gaudi 3 is cheaper + competitive. If Intel captures 15-20% of inference market, that’s $5B+ annual revenue by 2027.
3. Pat Gelsinger’s Bet Paying Off
CEO bet $50B on foundry pivot + AI chips. Wall Street doubted. This quarter proves he was right.
4. Cost Cuts Working
Opex down 12% YoY despite revenue +8%. Operating leverage = margin expansion ahead.
Risks
1. Valuation Undefined
INTC barely profitable (EPS $0.42/quarter = $1.68 annualized). P/E is ~60x forward. If growth slows, multiple compression is brutal.
2. Foundry Execution Risk
Ramping 18A to volume production = hard. TSMC has 20 years of experience. Intel has 2. Any yield issues = disaster.
3. NVDA Fights Back
If NVDA cuts inference pricing or launches competitive product, Intel’s Gaudi momentum stalls.
4. Macro Slowdown
If hyperscaler capex rolls over (recession, AI bubble pops), foundry orders evaporate.
Bottom Line
Best quarter in 3 years. Foundry ramp is real, AI chips are winning, PC business stabilizing. Stock dropped -2.5% despite 92% beat because it’s up 350% YTD — sell-the-news is natural.
Is the turnaround real? Yes. Foundry margins turning positive + hyperscaler design wins = structural change.
Is the stock still a buy? Not here. Wait for pullback to $85-90 range or buy on any macro dip. Valuation too stretched at $100+ after this run.
Next Catalyst: Q3 earnings in Oct. If foundry hits positive gross margin (as guided), stock re-rates higher. Watch for NVDA’s August earnings — if they warn on inference competition, Intel benefits.
Comp Check:
- AMD: P/E ~45 (vs INTC undefined)
- NVDA: P/E ~32 (but growing 80% YoY vs INTC 8%)
INTC is expensive on P/E but cheap on P/S if foundry story plays out. High-risk, high-reward from here.