Published: 2026-05-11 . Comparison: Intel (INTC) vs AMD . Price: INTC $129.97 / AMD $455.19 . Both ripped post-earnings (Intel +97%, AMD +28%)
One-Line Summary
"In the AI chip school, the #1 kid (NVIDIA) is too expensive, and the #3 / #4 kids (Intel, AMD) teamed up arguing 'AI inference needs CPUs too, not just GPUs.' Both aced the test, but the price tag got expensive."
Intel is going from "fading #1" to "reviving #1." AMD is locking in "AI #2." Both posted excellent earnings, but the stocks moved too fast, so the margin of safety is gone.
Market Cap and P/S Box
| Company | Per Share | Shares Out | Market Cap | Revenue | P/S |
|---|---|---|---|---|---|
| Intel | $130 | 4.83B | $628B | ~$54B | 11.5x 🔴 |
| AMD | $455 | 1.62B | $579B | ~$39B | 14.5x 🔴 |
Verdict (S&P avg 3x / semis normal 3-5x / cycle peak 5-6x / NVIDIA ~20x):
- INTC 11.5x is 2-3x above its 5-year normal range 🔴
- AMD 14.5x is ~2x above its 5-year normal range 🔴
Both look expensive, but AMD's 5x faster growth gives it a stronger case for its premium. Per-share price is just a function of shares outstanding.
What Do Intel and AMD Do?
Intel ($129.97)
One-liner: "The former #1 PC CPU maker, lost its way for a while, now reviving under a new CEO (Lip-Bu Tan)."
Three businesses:
- PC CPU (CCG): Laptop/desktop brains. 57% of revenue, ~$7-8B per quarter
- Server CPU + AI (DCAI): Cloud data center chips. 38% of revenue, +22% this quarter
- Foundry: Manufactures chips for other companies. First major customer: Google
Customers: MS, Amazon, Google, Meta and other hyperscalers plus all PC OEMs (Dell, HP, Lenovo).
AMD ($455.19)
One-liner: "The perennial #2 that CEO Lisa Su revived. Now the clear AI accelerator #2 behind NVIDIA, on the verge of crossing 50% server CPU share."
Four businesses:
- Data Center: EPYC server CPUs + Instinct AI accelerators. 57% of revenue, +57% this quarter
- Client (Ryzen): 28% of revenue, +26%
- Gaming: PlayStation/Xbox console chips + GPUs. +11% rebound
- Embedded: Xilinx-acquired industrial chips. +6% recovery
Customers: Meta (6GW mega-deal), OpenAI (6GW), Oracle, MS, console manufacturers.
How Big Are These Companies?
| Metric | Intel | AMD | Plain English |
|---|---|---|---|
| PC CPU share | 70%+ | 25%+ | Intel dominant in PCs |
| Server CPU share | 55% (falling) | 45% (rising) | Could flip within 2 years |
| AI accelerator share | <5% | ~10% | Both well behind NVIDIA (80%+) |
| Market cap | $628B | $579B | Both ~12x Korea's Kakao |
| Annual revenue | ~$54B | ~$39B | Intel larger |
Bigger Rivals (NVIDIA, Broadcom)
| Company | Position | Revenue Growth | GM | Take |
|---|---|---|---|---|
| 🏆 NVIDIA | AI accelerator king | +69% | 72% | The real #1 |
| 🥈 Broadcom | AI ASIC leader | +29% | 75% | Google TPU, Meta chips |
| 🥉 AMD | AI #2 + server CPU challenger | +38% | 55% | Catching up fast |
| #4 Intel | PC #1 + AI catch-up | +7% | 41% | Reviving but expensive |
Analogy: NVIDIA is the valedictorian. AMD is the student climbing from #3 to #2. Intel is the former #1 who slumped, got a new teacher, and is studying again.
What If These Companies Failed?
- If Intel failed: The US government would not let it happen for semiconductor security reasons (CHIPS Act). PC prices would spike.
- If AMD failed: NVIDIA monopoly, AI accelerator prices double. Console (PS5/Xbox) alternatives disappear.
Why Did Both Rip?
1. Intel: "Our 18A (next-gen process) is running ahead of internal plan"
New CEO Lip-Bu Tan: "18A wafers are running ahead of internal projections." This is the first signal that Intel is closing the gap on TSMC (#1 in Taiwan).
| Item | Last year | This time |
|---|---|---|
| 18A yield narrative | "concerns" | "ahead of plan" |
| Quarterly revenue | $12.67B | $13.58B |
| Operating margin | 5.4% | 12.3% 🟢 |
| Foundry anchor | none | Google secured 🟢 |
Analogy: A student who used to score 60 just got 85. The market is deciding "was that luck, or did they actually study?"
2. AMD: "AI Data Center +57%, Meta 6GW mega-deal"
Lisa Su: "Meta will deploy 6GW (= millions of units) of our Instinct GPUs." Combined with last year's OpenAI 6GW deal, AMD is now officially the AI accelerator #2.
| Item | Year-ago | This time |
|---|---|---|
| Data Center revenue | $3.7B | $5.8B (+57%) 🟢🟢 |
| Gross margin | 53.3% | 55% 🟢 |
| EPS | $0.96 | $1.37 (+43%) 🟢 |
Analogy: A basketball player who averaged 25 points a season now averages 35. The market accepts: "this is the new normal."
3. Shared Narrative: "AI inference also needs CPUs"
Both Intel and AMD CEOs delivered the same message:
"GPUs alone are inefficient for AI inference. CPUs need to run alongside. The GPU:CPU ratio is moving from 1:8 to 1:4, potentially toward parity."
If validated, both companies' server CPU TAM expands 2-4x. It's a direct counter to the NVIDIA / Broadcom solo narrative.
Analogy: People said "smartphones killed PCs," but the AI era is bringing PCs back.
4. Hyperscaler Capex +92%
Amazon, MS, Google, Meta, Oracle are putting roughly $750B into AI infrastructure in 2026, up +92% YoY. Both Intel and AMD benefit directly.
Big Picture: "Both are great, but expensive"
Which One Should You Buy?
| Comparison | Intel | AMD | Winner |
|---|---|---|---|
| Revenue growth | +7% | +38% | AMD |
| Gross margin | 41% | 55% | AMD |
| Q2 GM guide | 39% (down) | 56% (up) | AMD |
| Call tone (confidence) | 80% | 95% | AMD |
| Optionality (like 18A) | 🟢 yes | neutral | Intel |
| Mega-deals | Google x1 | Meta 6GW + OpenAI 6GW | AMD |
| Valuation | P/S 11.5x | P/S 14.5x | Intel slightly cheaper |
| Cycle position | peak imminent | peak imminent | Tie |
6 to 1 in favor of AMD. Intel's only edge is the 18A optionality.
So Why Aren't We Outright Buying?
1. Both ran too fast
- Intel: +97% post-earnings (D+18), 1Y +550%+
- AMD: +28% post-earnings, 1Y +380%+
- "Great company" does not equal "great stock price." 90-100% of the upside is already priced.
2. EV/EBITDA at cycle-peak levels
- Semis normal: 5-8x; cycle peak: 10-15x
- Intel 48-56x (4x+ peak) 🔴
- AMD 99x (~7x peak) 🔴🔴
3. First signs of hyperscaler FCF stress
- Amazon FCF guided -$17B (too much capex)
- Meta FCF -90% guide
- Implies a possible 'capex cut' within 1-2 years, which would slow semis revenue
4. Intrinsic models (SOTP) show downside
- Intel SOTP $108 vs spot $129.97 (-17%)
- AMD SOTP $332-396 vs spot $455 (-13% to -27%)
Verdict
| Ticker | 3M | 12M | $1,000 1Y | Allocation |
|---|---|---|---|---|
| AMD | Hold | Buy | ~$910 (-9%) | 55% |
| Intel | Sell | Hold | ~$800 (-20%) | 35% |
| (Cash) | 10% |
⚠️ Both face ~-17% downside on intrinsic models near-term. Conservative investors should consider diversifying into NVIDIA.
Are Fundamentals Actually Healthy?
Yes, the underlying numbers are strong.
| Item | Plain English | Intel | AMD |
|---|---|---|---|
| Revenue growth | Speed of getting bigger | 🟡 +7% | 🟢 +38% |
| Gross margin | Profit per $100 | 🟢 41% | 🟢 55% |
| Operating margin | Real operating profit % | 🟢 12.3% | 🟢 25% |
| EPS | Earnings per share | 🟢 $0.29 (back to profit) | 🟢 $1.37 (+43%) |
| GAAP loss | Accounting basis | 🔴 -$0.73 | 🟢 profit |
| FCF | Real cash generation | 🔴 -$3.9B | 🟢 healthy |
| Next-quarter GM | Margin trend | 🔴 -200bp | 🟢 +100bp |
| Call tone | CEO confidence | 🟢 recovering | 🟢 maxed |
Analogy: Both aced the test. AMD has 7 of 8 subjects at full marks; Intel has 5 full marks and 3 average grades.
Cycle Peak Check: Where Are We?
Semis are cyclical. 10 cycle-peak signals × 10 = thermometer. Above 70C = danger zone.
10-Signal Checklist
| # | Signal | Status | Score |
|---|---|---|---|
| 1 | OPM >50% | INTC 12% / AMD 25%, both below | ⬜ 0 |
| 2 | P/S >10x | INTC 11.5x / AMD 14.5x | ✅ 1 |
| 3 | EV/EBITDA >15x | INTC 48-56x / AMD 99x | ✅ 1 |
| 4 | Supply tight | TSMC full, Intel backlog | ✅ 1 |
| 5 | ASP rising | All raising prices | ✅ 1 |
| 6 | 1Y stock +200% | INTC +550% / AMD +380% | ✅ 1 |
| 7 | Revenue YoY +50% | Both below | ⬜ 0 |
| 8 | Big tech capex cuts | Still +92% raise, but Amazon FCF -$17B is first warning | ⬜ 0 |
| 9 | Spot prices stalling | Still rising | ⬜ 0 |
| 10 | China entrant pricing | Huawei ramping | 🟡 0.5 |
| Total | 5.5/10 |
Thermometer = 55°C "Early Boom"
| Temp | Stage | INTC est | AMD est | Action |
|---|---|---|---|---|
| 0-30°C | Ice age | $39-65 | $137-228 | 🟢 prime buy zone |
| 30-50°C | Recovery | $65-91 | $228-319 | 🟢 active buying |
| 50-70°C | Early boom ← now | $117-143 | $410-501 | 🟢 scale in OK |
| 70-85°C | Peak imminent | $130-169 | $455-592 | 🟡 start scaling out |
| 85-95°C | Peak passing | $150-182 | $523-637 | 🔴 reduce position |
| 95-100°C | Overheated | $169-195 | $592-683 | 🔴 sell now |
You can scale in here. If 1-2 more signals flip, we hit 65-75C "peak imminent" and new buys should stop.
The fastest-to-flip signal: #8 hyperscaler capex cut. Amazon FCF -$17B is the early warning.
So Buy or Sell?
Our View at a Glance
- Top Pick
- AMD
- AMD 3M
- Hold (scale in OK)
- AMD 12M
- Buy
- AMD Price Target
- $413
- Intel 3M
- Sell (turnaround overcooked)
- Intel 12M
- Hold
- Intel Price Target
- $104
- Conservative Alt
- Diversify with NVIDIA
Traffic-Light View
AMD
| Level | Action | Why |
|---|---|---|
| 🔴 $455-500 | No new buys | At top of 5Y P/S band |
| 🟡 $400-455 | Wait | Mean-revert zone |
| 🟢 $300-400 | Scale in OK | Base scenario attractive |
| 🛑 Below $290 | Reconsider | Bear scenario triggered |
Intel
| Level | Action | Why |
|---|---|---|
| 🔴 $130-150 | Sell / trim | Turnaround 100% priced |
| 🟡 $100-130 | Wait | Near fair value |
| 🟢 $75-100 | Scale in | 18A optionality attractive |
| 🛑 Below $60 | Reconsider | 18A failure scenario |
$1,000 Investment (1 Year)
AMD
| Scenario | Probability | 1Y |
|---|---|---|
| 🟢 Bull | 25% | $1,230 (+23%) |
| 🟡 Base | 50% | $900 (-10%) |
| 🔴 Bear | 25% | $640 (-36%) |
| Expected | ~$910 (-9%) |
Intel
| Scenario | Probability | 1Y |
|---|---|---|
| 🟢 Bull | 20% | $1,120 (+12%) |
| 🟡 Base | 50% | $830 (-17%) |
| 🔴 Bear | 30% | $580 (-42%) |
| Expected | ~$800 (-20%) |
⚠️ Both have negative 12M expected values. Cycle-peak-level valuation is the cause. Diversifying into NVIDIA or Broadcom may be the smarter call.
What to Watch Next? (5-Item Checklist)
Next earnings: Intel late July, AMD early August.
| Priority | What | Bull | Bear |
|---|---|---|---|
| 🥇 | Big tech capex guides | further raise | first cut → reduce immediately |
| 🥈 | GPU/CPU ASP trend | rising | stalling or declining |
| 🥉 | Days inventory | <30 days | >45 days |
| 4 | (Intel) 18A yield disclosure | 70%+ | undisclosed or below 50% |
| 5 | (AMD) MI350 adoption list | more names beyond Meta | only 1-2 |
3+ green = upgrade AMD to stronger BUY. 3+ red = downgrade both PTs.
FAQ
Q1. AMD at $455 feels expensive. Why not just buy the cheaper $130 Intel? A. Per-share price is a trap. On P/S, Intel 11.5x vs AMD 14.5x is only a 25% gap, while AMD grows 5x faster. AMD's premium is more justifiable.
Q2. NVIDIA is leading, why not just buy NVIDIA? A. This analysis leans that way. NVIDIA revenue +69%, GM 72%, P/S ~20x, more reasonable valuation. For diversification, something like NVDA 60% / AMD 30% / INTC 10% is safer.
Q3. Is Intel's "18A is going well" claim credible? A. No quantitative yield was disclosed. If a real number drops in the next 1-2 quarters, +20% gap up is possible; a disappointment means -30%. Pure optionality.
Q4. When does the Meta 6GW deal hit revenue? A. Usually 18-24 months from LOI to shipment to recognition. The bulk lands in 2027-28. But the CEO mentioned "specific data center locations identified," which implies unusually strong visibility.
Q5. If cycle peak is near, should I just sell everything? A. We are at 55C, "early boom," not the peak. Above 70C we start scaling out. The most important sell trigger is signal #8 (hyperscaler capex cut).
Risk Warnings
⚠️ Both Intel and AMD are high-volatility names:
- Daily moves of ±3-5% are normal
- ±20% moves possible at the July/August earnings
- Putting in $1,000? Be prepared for -30% short-term moves
- A -50% scenario exists if a downcycle starts right after the peak
⚠️ Watch for these scenarios:
- Amazon, MS, or Meta cuts capex guidance → -20% immediate move possible
- NVIDIA Blackwell Ultra price aggression → AMD MI350 ASP pressure
- New China export controls → AMD -10% immediate, INTC indirect impact
⚠️ This is not investment advice. Decide based on your own judgment.
Jargon Dictionary
| Term | Plain |
|---|---|
| P/S | Market cap / revenue. "Price per $1 of sales" |
| EV/EBITDA | Enterprise value / operating cash. Better than P/E for semis |
| SOTP | Sum-of-the-parts; valuing each segment separately. Required because no segment exceeds 70% of revenue |
| GM | Gross margin; profit per $100 of sales |
| OPM | Operating margin |
| FCF | Operating cash minus capex (real cash) |
| CapEx | Capital expenditure (factories, equipment) |
| Hyperscaler | Amazon, MS, Google, Meta, Oracle |
| 18A | Intel's 1.8nm next-gen process. Competing with TSMC N2 |
| EPYC | AMD server CPU brand |
| Instinct (MI350) | AMD AI accelerator series |
| DCAI | Intel Data Center & AI segment |
| CCG | Intel Client Computing (PC CPU) |
| Inference | Using a trained AI model. CPU-intensive vs training |
| Cycle peak | The high point of an industry cycle; usually followed by a drop |
| CapEx cut | When big tech starts trimming investment, a peak signal |
For the deeper data, SOTP build, and cycle scorecard, see the Intel vs AMD analyst comparison. Same earnings season: Cloudflare -24% crash, Datadog +30% surge, SanDisk beginner guide. New to stocks entirely? Start with the Stock Investing Beginner Guide.
This report is for educational analysis only and is not investment advice. US stocks also move on USD/KRW FX; keep that in mind.