Skip to content

🐂 NVDA — Company Profile (Teaching)

Teaching site, not investment advice. Specific market cap / stock price / earnings figures change frequently and are intentionally omitted. Use SEC + Yahoo Finance links below for real-time data.

📌 Basics

Dimension Content Source
Core business GPU + CUDA ecosystem + Mellanox networking (AI accelerator + data center fabric) 10-K FY25
Founded / HQ 1993 / Santa Clara, California, USA wiki
CEO Jensen Huang (since founding 1993, founder-CEO) Proxy 2026
Primary listing NASDAQ: NVDA —
Fiscal year 52/53-week year ending last Sunday of January (FY26 = 2026/01/25; FY27 Q1 = late Jan – late Apr 2026; earnings released 5/20) SEC EDGAR

💡 Real-time market cap / share price / latest quarterly revenue / GM: see SEC primary + IR links below

🏭 Supply Chain Coordinates (Part 1-C6 framework)

Upstream (Who NVDA depends on)

Supplier What flows Key dependency
TSM GPU foundry (CoWoS packaging) + leading-edge nodes Irreplaceable (Samsung foundry / Intel foundry 1-2 nodes behind)
SKH (KRX 000660) HBM3E primary supplier (70%+ share) Strong lock-in post-qualification
MU HBM3E 2nd supplier Backup capacity
SNPS + CDNS EDA software (chip design tools) Duopoly, physical dependency
COHR + LITE Optical modules (800G → 1.6T → CPO) NVDA invested $2B in COHR + $2B in LITE (combined $4B, announced 2026/03); agreements non-exclusive, with multibillion purchase commitments + future capacity / access rights

Downstream (Who depends on NVDA)

Customer Relationship Customer concentration
MSFT #1 hyperscaler customer Top 4 hyperscalers account for ~40%+ of NVDA data center revenue (per official 10-K customer concentration disclosure)
META #2 hyperscaler
GOOGL Hyperscaler + partial self-supply via TPU
AMZN Hyperscaler + partial self-supply via Trainium
ORCL OCI + Stargate (newly major customer; reported $300B / 5-yr OpenAI contract)
CRWV Neocloud king (250K+ GPU; NVDA strategic stake)
OPENAI Direct strategic partner: NVDA 10 GW deployment agreement

Competitors

  • Direct (independent GPU): AMD (MI300 / 350; but ROCm software ecosystem weak, market share 5-10%)
  • Customer self-developed ASIC: GOOGL TPU / AMZN Trainium / MSFT Maia (mainly internal workloads)
  • China domestic substitutes: Huawei Ascend 910C (China internal market, not sold overseas)
  • Co-design: AVGO (Broadcom co-designing 10 GW custom ASIC with OpenAI)

💼 Business Model + Value Capture (Part 1-C7 view)

  • Revenue mix: Data center ~90%+ (mainly hyperscalers + neocloud), gaming ~5%, automotive + professional visualization + other ~3%
  • Gross margin: 70%+ (among historical highs, supported by physical + ecosystem dual moat)
  • Pricing power: H100 → H200 → Blackwell price jumps 30-50% per generation; customers locked in by CUDA, hard to refuse
  • Substitution difficulty: Extremely high — replacers must simultaneously catch up on CUDA ecosystem (20-year accumulation) + Mellanox networking + hyperscaler strategic relationships, 3 layers each requiring 5-10 years, near-impossible to do simultaneously

(For specific quarterly revenue / GM / segment numbers, see latest SEC 10-Q)

🏰 Moat Assessment (Part 3-C2 Buffett 5-step lens)

Type Strength (0-1) Source
Brand 0.8 "NVIDIA" is virtually synonymous with "AI compute"
Switching cost 1.0 CUDA ecosystem 20 years, global university curriculum base, AMD ROCm still can't catch up
Network effects 0.7 Developer + university + ISV (Adobe / Autodesk / various ML frameworks) ecosystem
Scale + cost 0.8 TSM priority capacity + in-house NVLink + massive R&D investment

Composite: 3.3 / 4 → extremely strong (tech + ecosystem dual moat)

Note: Part 4-C5 uses strict Buffett lens and rates NVDA's moat only 2/4 (because 10-year visibility insufficient + valuation too high = outside circle). Two scores using different lenses don't conflict.

⚠️ Key Reverse-Trigger Conditions (Part 3-C5 anti-thesis view)

Monitor the following signals; if triggered, re-evaluate thesis:

  • Trigger 1 (quarterly): Top 4 hyperscalers' combined capex YoY growth < 10% (vs historical 25%+) — monitor each company's FY guide
  • Trigger 2 (quarterly): OpenAI valuation reset down-round or revenue growth < 50% YoY
  • Trigger 3 (event): 1 or more hyperscalers publicly states "we've over-invested in AI capex"
  • Trigger 4 (annual): Hyperscaler self-developed ASIC (TPU / Trainium / Maia) internal share > 50% — long-term share-squeeze signal
  • Trigger 5 (low probability, catastrophic): Taiwan Strait event → TSM shipment disruption → NVDA full chain halt 6-12 months

🔗 Teaching Cross-References

NVDA is the most referenced ticker on-site, covered in many chapters:

📚 SEC Primary Public Filings (DYOR)

Real-time quote: Yahoo Finance NVDA

Teaching use · numbers change — learn the method, not the specific figures · investment decisions are based on your own research + risk tolerance