earnings
🟢 Q1 EPS $1.61 vs $1.42 consensus; Revenue $11.27B vs $10.71B consensus, both beating on every line. (high conf)
guidance
🟢 Q2 guide raised to $12.5B vs $11.2B consensus; Data center revenue $7.2B (+65% YoY); MI325 GPU ramping with estimated $1-1.5B incremental contribution. (high conf)
product
🟢 CEO Lisa Su cited CPU demand doubling, indicating AMD shifting from GPU-only to CPU+GPU dual-track positioning; MI300X hyperscaler dual-sourcing strategy continuing to pay off. (high conf)
competition
🟢 AMD being redefined from 'challenger' to 'hyperscaler indispensable second source'; Q2 actual data center revenue above $7.5B would confirm sustainable acceleration; below $6.8B would suggest channel stuffing. (medium conf)
guidance
🟢 FY27 data center TAM revised upward from $35-38B to $42-45B consensus; server CPU TAM expected to double to $120B+ by 2030 as CPU:GPU ratio shifts from 1:4 to 1:1 for agentic AI workloads. (high conf)
Thesis Impact
Reinforces growth thesis — AMD guide $12.5B significantly above consensus validates AI infrastructure capex acceleration; Q2 data center revenue above $7.5B will confirm sustained multi-quarter acceleration.
supply_chain
🟢 Nvidia announced up to $3.2B investment in Corning to build three new factories in North Carolina and Texas dedicated to AI datacenter optical components; aims to diversify away from single-supplier dependency on II-VI/Coherent and Lumentum. (high conf)
product
🟢 Vera standalone CPU designed for AI-factory orchestration could capture 20-25% of server CPU market; open-sourced cuDNN kernels (MoE, sparse attention) accelerate developer innovation on Nvidia hardware. (medium conf)
earnings
⚪ Nvidia flat on day while AMD +19%, Intel +5%, SOXX at all-time high — significant relative performance underperformance despite strong sector; May 20 earnings expected to resolve whether stock has supply issues or valuation-based stall. (medium conf)
macro
🔴 GH200 supply constraint reported to be easing, which could compress AMD's window as second-source; Nvidia's supply situation vs competitors' momentum is key near-term catalyst. (medium conf)
Thesis Impact
Challenges existing thesis — flat performance while AI peers rally suggests potential valuation cap or unrecognized supply concerns; May 20 earnings report is critical catalyst to resolve this divergence.
product
🟢 First disclosure of $2B+ customer demand (FY27-28 combined) covering first proprietary datacenter CPU; transition from IP licensing to chip designer is a major milestone. (high conf)
earnings
⚪ Stock initially surged then pulled back as market interpreted $2B as design-win pipeline, not revenue; actual revenue conversion earliest FY28; FY27 conversion rate will be key data point. (high conf)
capex
⚪ Design-win pipeline requires 12-18 months to convert to revenue; market currently pricing datacenter CPU business at high multiple; FY27 actual revenue miss would compress multiple materially. (medium conf)
Thesis Impact
Reinforces thesis — $2B+ pipeline validates datacenter CPU entry but 12-18 month revenue lag and sell-the-news reaction suggest risk-reward at current levels needs recalibration.
earnings
🟢 Fortinet +24% in single session — largest billings expansion in three years, representing significant acceleration in network security demand. (high conf)
macro
🟢 AI infrastructure buildout driving incremental network security spending; billings expansion may signal broader enterprise capex acceptance in security sector. (medium conf)
Thesis Impact
Reinforces thesis — largest billings expansion in three years signals enterprise security spending acceleration tied to AI infrastructure buildout, validating broader sector exposure.
earnings
🟢 Intel +5% on the day as capital rotated from GPU-premium names to CPU/memory-oriented positions within semiconductor sector. (medium conf)
macro
⚪ Internal rotation within SOXX from GPU-heavy to CPU/memory direction signals AI infrastructure行情 expanding beyond single-core (Nvidia) to multi-core (multiple sub-sectors). (medium conf)
Thesis Impact
No thesis impact — +5% reflects sector rotation dynamics rather than Intel-specific catalyst; limited actionable signal without earnings confirmation.
supply_chain
🟢 Micron continued strong post-earnings; memory cycle confirmed by LTA shift (Samsung and SK Hynix moving to 3-5 year contracts), NAND Flash contract prices rising, SSD business at all-time high. (high conf)
earnings
🟢 No cycle top signals yet; LTA willingness from suppliers signals belief in sustained shortage rather than cycle peak cash-out behavior — pattern structurally different from 2018 memory cycle. (high conf)
product
🟢 HBM remains binding constraint; Samsung hit $1T market cap, SK Hynix surging on memory boom; DDR5 remains tight; 7th-gen DRAM innovations signal structural undersupply for AI compute applications. (high conf)
Thesis Impact
Reinforces thesis — LTA shift confirms structural (not cyclical) memory demand driven by AI; shortage depth narrative remains intact with no cycle-top signals present.
supply_chain
🟢 Nvidia committed up to $3.2B investment for Corning to build three AI datacenter optical component factories in North Carolina and Texas; represents AI optical infrastructure official certification. (high conf)
product
🟢 Optical network is frequently-ignored AI infrastructure bottleneck; new optical factory requires 18-24 months vs GPU fab cycle; Corning securing this deal provides clear order visibility in a supply-constrained segment. (medium conf)
capex
🟢 Stock +5% on announcement, +8% intraday high; optical business currently ~15-18% of revenue; Nvidia deal may prompt management to give clearer AI business segment disclosure at next CFO transition. (medium conf)
Thesis Impact
Reinforces thesis — Nvidia partnership is structural catalyst, not transactional; 18-24 month fab cycle means optical component supply will remain tighter than GPU longer, providing sustained order visibility.
earnings
🟢 Disney +7%; Q2 streaming + parks revenue beat — first quarterly report under Josh D'Amaro showing clean revenue beat; streaming business entering sustainable profitability phase. (high conf)
macro
🟢 Premium consumer spending ($80k+ income households) showing resilience; Disney and Uber rallies reflecting same consumer premiumization thesis. (medium conf)
Thesis Impact
Reinforces thesis — streaming profitability milestone validates transformation narrative; premium consumer resilience supports continued parks/streaming growth trajectory.
earnings
🟢 Uber +9%; Q2 bookings beat +9%, Uber One membership exceeded 50 million, trips and gross bookings both grew 20%+. (high conf)
macro
🟢 Premium consumer spending resilience confirmed; high-income household spending not slowing, supporting gig economy and experiential consumption thesis. (medium conf)
Thesis Impact
Reinforces thesis — 20%+ trip growth and Uber One milestone validate consumer premiumization trend; strong engagement metrics support continued margin expansion narrative.
guidance
🟢 DoorDash +12%; gross order value guidance beat; revenue miss ignored by market — market showing high tolerance for 'growth with improving margin' over pure revenue beat. (high conf)
product
🟢 Margin expansion timeline in next earnings report would be next catalyst; market currently pricing DoorDash on path to profitability milestone, not current earnings. (medium conf)
Thesis Impact
Reinforces thesis — market preference shifting to 'growth with improving margin'; GOV beat validates unit economics strength; margin expansion timeline would trigger further re-rating.
earnings
🟢 Google Cloud +63% YoY, partly fueled by Anthropic commitment; 1Q26 capex $35.7B with FY guide $180-190B. (high conf)
product
🟢 Anthropic committed $200B to Google Cloud over 5 years, representing >40% of Google's backlog; combined AI lab contracts now exceed 50% of total hyperscaler backlog. (high conf)
capex
🟢 2026E capex tracking $180-190B; hyperscaler total capex now ~$700-725B industry-wide; cloud AI monetization improving as Anthropic added ~$10B annualized run-rate in 30 days. (high conf)
Thesis Impact
Reinforces thesis — Anthropic $200B commitment validates Google Cloud AI demand visibility; cloud +63% YoY confirms competitive positioning; AI monetization improving sustainability of capex.
earnings
🟢 Azure growth re-accelerating into 40s for 2H26; Commercial Office accelerating suggesting Copilot user conversion improving marginal revenue/capex; 1Q26 capex $30.9B with FY guide $190B. (high conf)
capex
🟢 Full-year capex target $190B; Azure deceleration reversal suggests AI monetization pathway working; Copilot conversion metrics improving. (medium conf)
Thesis Impact
Reinforces thesis — Azure re-acceleration confirms AI monetization working; Copilot conversion improving validates productivity narrative; FY capex at $190B confirms sustained AI infrastructure commitment.
capex
🟢 Amazon targeting $200B full-year capex 2026; 1Q26 combined with AWS AI buildout; hyperscaler capex cycle sustaining multi-year demand visibility. (high conf)
macro
⚪ Hyperscaler combined 2026E capex ~$700-725B up from $670B; concentration risk noted as hyperscalers now ~20% of IG Dv01 issuance. (high conf)
Thesis Impact
Reinforces thesis — $200B capex target confirms AWS AI infrastructure commitment; hyperscaler capex supercycle sustaining 2-3 year demand visibility for AI infrastructure suppliers.
capex
🟢 Meta 2026E capex $125-145B; arranged $13B financing (mostly debt) for El Paso 1GW datacenter by 2028, following $27B SPV for Louisiana; hyperscalers increasingly using off-balance-sheet structures. (high conf)
regulation
🟢 Meta is Trump admin ally; Scale AI (Meta-backed) Pentagon contract signals political access advantage for B2G market; commercial AI companies face structural ceiling in government sector. (medium conf)
product
🟢 Building 'Hatch' AI agent for Instagram shopping; consumer agent deployment widening training/inference compute demand. (medium conf)
Thesis Impact
Reinforces thesis — Pentagon contract access (via Scale AI) confirms political moat forming; $13B financing shows capital markets accommodating AI buildout; consumer AI agents add new compute demand vector.
capex
🟢 TSMC investing NT$100B ($3.1B) in Fab 15A upgrade to 4nm and 1.4nm; older tools relocating to Dresden; demonstrates flexible capacity reallocation toward AI leading-edge. (high conf)
supply_chain
🟢 GlobalWafers raising 12-inch wafer prices on AI demand; Sigurd Micro raising capex 49% to NT$8.8B for packaging/testing — direct beneficiaries include Advantest and BESI; confirms accelerating back-end investment. (high conf)
product
⚪ China pushing for 70% domestic 12-inch wafer production by year-end — long-term substitution risk but near-term limited given AI demand surge. (medium conf)
Thesis Impact
Reinforces thesis — TSMC capacity upgrade and wafer price hikes confirm AI leading-edge demand sustained; back-end capex acceleration signals full supply chain benefiting from AI infrastructure buildout.
product
🟢 Palantir's 'AI slop' framing validated — 10-step agent chain with 95% step reliability yields only ~60% end-to-end accuracy; lower cost shifts bottleneck from price to compounding errors; governance/ontology becoming structural moat. (medium conf)
product
🟢 Palantir agentic workflows accelerating token consumption — tangible enterprise AI scale; 'no slop zone' positioned as competitive advantage for mission-critical applications. (medium conf)
Thesis Impact
Reinforces thesis — accuracy compounding problem validates Palantir's quality-over-speed approach; governance moat increasingly relevant as AI deployment scales; enterprise value proposition strengthened.