AMD Q1 2026 — Earnings Synthesis & Updated Price Target
One-line: Lisa Su effectively rewrote the financial model — doubled the server-CPU TAM, added an explicit "$20+ EPS over the strategic time frame" anchor. The Tarot RSU hurdles use a 90-day trailing average vesting mechanic, so at spot $458 (90-day avg $242) the in-the-money tranches Pink's prior analysis flagged at $397 are still trajectory-dependent, not locked.
Revision note (2026-05-12 PM): This synthesis was rewritten after pulling Pink's claude.ai chat
AMD share price tranche calculationsvia/claude-reader. The prior version under-specified the 8-K mechanics. Corrected facts now in §3.
Live data (as of 2026-05-12 18:57 ET):
- Spot: $458.79 | 52w range $108–$469 | market cap $748B | shares 1.63B | fwd P/E 36x | trailing P/E 154x
- 30-day avg close: $299 | 60-day: $250 | 90-day: $242 ← this is the hurdle measurement
1. What changed vs. prior guidance
| Metric | Prior (Nov 2025 Analyst Day / Mar 2026 MS conf) | New (Q1 2026 call, 5 May 2026) | Delta |
|---|---|---|---|
| Server CPU TAM (2030) | $60B at 18% CAGR | $120B at >35% CAGR | 2x |
| Long-term EPS target | Not explicit | ">$20 EPS over strategic time frame" | new anchor |
| Long-term gross margin | "Expanding" | 55–58% range (Jean Hu) | quantified |
| Data Center 3–5yr growth | >60% annual | Restated; reiterated track to "exceed long-term targets" | reaffirmed |
| Instinct/AI revenue | "Tens of billions by 2027" | "Strong and increasing confidence" in 2027 | confidence step-up |
| Instinct long-term growth | Not specified | ">80% to exceed in coming years" | new floor |
| AMD server share aspiration | Implied parity with Intel | ">50% share" of expanded $120B TAM | explicit |
Key Lisa Su quotes (verbatim from transcript)
"we now expect the server CPU TAM to grow at greater than 35% annually, reaching over $120 billion by 2030."
"Agentic AI and inferencing and all the CPU compute that is required, is just happening, and it's happening at a much faster pace."
"we have strong and increasing confidence in our ability to deliver tens of billions of dollars in annual Data Center AI revenue in 2027"
"deliver more than $20 in EPS over the strategic time frame"
"exceed our long-term growth target of greater than 80% [Instinct] in the coming years"
This is the same Lisa who at Morgan Stanley in March said demand "far exceeded my expectations." Q1 is the operational confirmation of that.
2. Q1 2026 actuals & Q2 guide
| Q1 2026A | Q2 2026E | YoY | |
|---|---|---|---|
| Revenue | $10.3B | $11.2B (±$300M) | +38% / +46% |
| Data Center | $5.8B | — | +57% |
| DC operating income | $1.6B (28% margin) | — | +72% |
| Server CPU YoY | "+50%" | "+70%" | accelerating |
| EPYC cloud instances | 1,600+ | — | +50% YoY |
| Non-GAAP GM | — | 56% | within 55–58% LT band |
| OpEx (non-GAAP) | — | $3.3B |
Note the acceleration in server CPU growth from +50% Q1 to +70% guided Q2 — this is what "much faster pace" means in numbers. Supply is now allocated; this is order-book reality, not projection.
3. The RSU package — corrected mechanics (from 8-K + claude.ai chat reconciliation)
2026-04-26-2026-grants-digest-volume-5 flagged the 17-Feb-2026 filing: a special $75M CEO Value Creation Equity Award for Lisa Su, granted 15-Mar-2026, performance period through 15-Mar-2031 (5 years). The Tarot summary was directionally right but the 8-K has critical mechanic details that change the live picture.
8-K mechanics (verified):
- Baseline = 60-trading-day average closing price ending Grant Date (15-Mar-2026). Effective baseline ≈ $203 — Lisa's own 10b5-1 sales on 12-Mar were $197.42–$203.15, and the 8-K notes the Committee considered AMD's all-time high of $267.08 as of grant date when designing hurdles.
- Vesting = the 90-trading-day trailing average closing price must meet/exceed each hurdle during the performance period. It is sustained, not single-day-touch.
- T1, T2, T3 are FIXED dollar values set on Grant Date, rounded to nearest $5. T4 is absolute $600.
- Vesting starts year 3; full payout year 5; tenure required throughout.
Four tranches (at $203 working baseline, rounded to nearest $5):
| Tranche | Payout | Structure | Hurdle (90d avg) |
|---|---|---|---|
| T1 | 50% | 10% 5yr CAGR | ~$325 (+60%) |
| T2 | 100% | 15% 5yr CAGR | ~$410 (+102%) |
| T3 | 150% | 17.5% 5yr CAGR | ~$455 (+124%) |
| T4 | 200% | Absolute | $600 (~3x; ~$1T mkt cap) |
Tranche status as of 2026-05-12 (spot $458, 90-day avg $242)
| Tranche | Hurdle | 90-day avg today | Spot today | Status |
|---|---|---|---|---|
| T1 ~$325 | $325 | $242 | $458 | Spot through; 90d avg ~25% below. If AMD holds $458, 90d avg crosses $325 ~Jun-Jul 2026 |
| T2 ~$410 | $410 | $242 | $458 | Spot through; 90d avg ~40% below. 90d avg crosses ~Aug-Sep 2026 if AMD holds |
| T3 ~$455 | $455 | $242 | $458 | Spot at hurdle. 90d avg needs sustained ~$455+, which means stock has to hold ~$455+ from here for 90+ days |
| T4 $600 | $600 | $242 | $458 | Needs ~31% more spot, then 90 days of sustaining |
The May-5 claude.ai chat said "T1/T2 are functionally won at $397." That framing reads spot vs hurdle. The 8-K vesting mechanic is stricter — the 90-day trailing average must reach the hurdle. At today's $458 the 90-day avg is still $242 (because most of the 90-day window was below $300). T1/T2 are on the trajectory to win in Q3-2026 if the stock holds. They are NOT yet won.
4. Why the RSU hurdles are calibrated to Lisa's newly raised model
Build a base-case EPS bridge using the new model (28% op margin, 14% tax, 1.63B basic / ~1.66B diluted shares per latest filings):
| Year | Revenue | Op income | Net income | EPS |
|---|---|---|---|---|
| FY25A | $34.6B | $7.6B | $6.5B | $3.93 |
| FY26E (Q1+38, Q2+46 → ~+40% full year) | $48B | $13.6B | $11.7B | $7.0 |
| FY27E (+35%, AI ramps) | $65B | $18.3B | $15.7B | $9.5 |
| FY28E (+30%, MI450 mature) | $85B | $23.8B | $20.5B | $12.3 |
| FY29E (+22%) | $104B | $29.0B | $25.0B | $15.0 |
| FY30E (+25%, op-margin ratchet to 30%) | $130B | $39B | $33B | $20.0 |
→ $20+ EPS is reachable by FY30 at sustained 30%-ish CAGR with op-margin expansion to 30%.
Cross-check (Pink's claude.ai chat, Apr 27): the analyst-day bottom-up math is more aggressive. Data center >60% CAGR + core >10% CAGR on 2025 mix of 48% / 52% → ~$203B revenue by 2030. At 28% op margin, 14% tax, 1.66B shares → ~$31 EPS by FY30. So the $20 EPS anchor is closer to the low end of Lisa's own analyst-day floor scenarios, not the midpoint.
Now apply the multiple at each tranche:
| Tranche | Hurdle price | Multiple at $20 EPS | Multiple at $15 EPS | Multiple at $31 EPS (analyst-day bottom-up) |
|---|---|---|---|---|
| T1 | $325 | 16x | 22x | 10x |
| T2 | $410 | 21x | 27x | 13x |
| T3 | $455 | 23x | 30x | 15x |
| T4 | $600 | 30x | 40x | 19x |
The math: at $20 EPS (Lisa's stated anchor) × 30x (current forward P/E) = $600 = T4 exactly. The board calibrated the maximum tranche to Lisa's stated floor scenario. If she delivers the analyst-day bottom-up (~$31 EPS), T4 vests at a 19x multiple — well below today's 36x forward.
The board signed off on a comp package whose top tranche pays out only if Lisa delivers her own newly raised long-term financial model. She announced the model on the Q1 call seven weeks after the grant date. The "RSU as insider-confidence signal" thesis from the Tarot post is intact — and the post-Q1 raise is the public follow-through.
5. Updated price target — synthesis (re-anchored to spot $458)
Current spot $458.79. Pink's actual cost basis: needs confirmation — Decision Log pre-buy analysis was $192.43 but that is a recommendation date, not a fill. Return math below uses $458 as the decision anchor.
Forward-PE reality check: AMD trades at 36x forward today. That's ~50% above its 5-year median, and prices in much of Lisa's raised model already.
12-month forward (May 2027), anchor FY27 EPS $9.5
| Case | EPS | Multiple | Target | vs $458 |
|---|---|---|---|---|
| Bear (AI GPU slips, multiple compresses to AI-comp lows) | $8 | 25x | $200 | −56% |
| Base | $9.5 | 32x | $304 | −34% |
| Multiple holds | $9.5 | 40x | $380 | −17% |
| Bull (multiple expands to NVDA-comp 50x) | $10 | 50x | $500 | +9% |
→ At $458, the 12-month risk/reward is asymmetric to the downside unless the AI re-rate continues and Q2 confirms +46% guide.
24-month forward (May 2028), anchor FY28 EPS $12.3
| Case | EPS | Multiple | Target | vs $458 |
|---|---|---|---|---|
| Bear | $10 | 25x | $250 | −45% |
| Base | $12.3 | 32x | $394 | −14% |
| Bull | $14 | 40x | $560 | +22% |
5-year forward (Mar 2031, RSU horizon), anchor Lisa's $20+ EPS
| Case | EPS | Multiple | Target | vs $458 | CAGR | RSU vests |
|---|---|---|---|---|---|---|
| Miss (60% of model) | $12 | 20x | $240 | −48% | −13%/yr | none |
| Bear (75% of model) | $15 | 22x | $330 | −28% | −6%/yr | T1 only |
| Base ($20 anchor hit) | $20 | 25x | $500 | +9% | +1.7%/yr | T1+T2+T3 |
| Bull (analyst-day bottom-up) | $25 | 28x | $700 | +53% | +9%/yr | T1+T2+T3+T4 |
| Stretch (full bottom-up + AI re-rate) | $31 | 30x | $930 | +103% | +15%/yr | T1+T2+T3+T4 |
Recommended PTs for Pink's book
- 12-month PT: $400 (32x FY27 $9.5 with modest multiple holding; ~$60 below spot — i.e., "the stock has gotten ahead of FY27 EPS")
- 5-year PT: $600 (= T4 = Lisa's $20 × 30x = the explicit board ceiling; +31% / +5.6% CAGR)
- Stretch 5yr: $750–900 if analyst-day bottom-up scenario plays out
What this means at $458
The market has compressed 18-24 months of expected appreciation into ~6 weeks. Three honest reads:
- The easy money was made. From the $192 wiki entry (Mar 9) to $458 today (May 12) is +139% in 9 weeks. That's the post-Q1 rerate.
- Forward returns are now market-like. Base case 5yr CAGR ~1.7% to PT $500. Bull case ~9%. Need the analyst-day bottom-up scenario to get a 15% CAGR — which is doable, but it's the bull case, not base.
- Mostly a Tranche 4 option from here. T1/T2/T3 90-day-avg vesting happens almost automatically if AMD holds these levels. T4 ($600) is the live question and ~$140 of further upside.
Action framework
| Conviction level | Action |
|---|---|
| High (you believe analyst-day bottom-up) | Hold full. Add only on pullback to $350 (~24x FY27 EPS, where the post-Q1 rerate started) |
| Medium (you believe Lisa's $20 EPS anchor) | Trim 25–33% of position here at $458. Keep the rest for T4. The trim locks the post-Q1 windfall; the remainder is the Lisa-comp-aligned 5-year option |
| Low (you think AI is froth) | Trim 50%+. AMD at 36x forward is expensive, and 12M math says downside if Q2 doesn't confirm |
| Add | Not at $458. Wait for a pullback to $350–380, or for Q2 confirmation in early August |
6. Why I'm raising the PT
- The TAM doubled. $60B → $120B server CPU TAM by 2030 isn't a marginal raise; it's a step-function in the addressable market for AMD's strongest franchise.
- An explicit EPS anchor exists now. "$20+" is what management is being paid to deliver. Aligning the model to that anchor is the rational base case for an outside investor.
- Q2 server CPU +70% is order-book reality. Supply allocated through FY26; lead times 8–10 weeks. The growth is contracted, not projected.
- AI revenue inflection moved up. "Strong and increasing confidence" in 2027 tens-of-billions = MI400/MI450 customer commitments are in hand, not pipeline.
- The RSU comp is the strongest insider-confidence signal of 2026. Lisa already holds ~$765M in stock; the new $75M grant only pays out on the trajectory she just publicly committed to.
7. Risks unchanged from base thesis
- NVIDIA CUDA moat — primary cap on Instinct upside; if 80%+ growth target slips it knocks ~$5 off the FY30 EPS bridge → PT down to ~$400 stretch.
- NVIDIA Vera CPU — direct attack on the CPU TAM thesis. Not shipping yet but Meta is committed.
- TSMC 2nm capacity — if Venice is supply-constrained, FY27 misses.
- China export curbs widening — $1.5B+ revenue at risk.
- Valuation already rich — 30x forward; any execution stumble re-rates fast. The -17% post-Q4-2025 reaction is the playbook.
8. What to watch next
- November 2026 Financial Analyst Day — does Lisa formalize the $120B TAM and $20+ EPS into multi-year financial framework?
- Q3 2026 MI450/Helios initial shipments — Q4 is the "significant ramp." Misses here break the AI inflection.
- Server CPU share data — Mercury Research quarterly prints. Track toward the ">50% of $120B" aspiration.
- Vera launch timing — biggest competitive risk; track NVIDIA earnings commentary.
Sources
- AMD Q1 2026 Earnings Call Transcript — Motley Fool, 6 May 2026
- AMD Q1 2026 Earnings — Investing.com
- AMD Q1 2026 Earnings — Seeking Alpha
- AMD Q1 2026 Earnings Call Presentation — Seeking Alpha
- 2026-04-26-2026-grants-digest-volume-5 — Tarot Capital RSU breakdown (paid)
- AMD — canonical wiki page
- amd-filings — filings tracker