Skip to content

🐂 ANET — Company Profile (Teaching)

Teaching site, not investment advice. Specific market cap / stock price / financial figures change frequently and are intentionally omitted; use the SEC + Yahoo Finance links below to pull real-time data yourself.

📌 Basics

Dimension Content Source
Core business Datacenter Ethernet switches (high-speed 100G/400G/800G) + EOS operating system + Etherlink AI network fabric — a leading AI cluster network option (vs. NVDA Mellanox InfiniBand) 10-K FY25
Founded / HQ 2004 / Santa Clara, California, USA wiki
CEO Jayshree Ullal (appointed President & CEO 2008/10/23; not a founder — co-founders are Andy Bechtolsheim / David Cheriton / Ken Duda; long-tenured) DEF 14A
Primary listing NYSE: ANET
Fiscal year Calendar year (Jan 1 – Dec 31) SEC EDGAR

💡 Real-time market cap / share price / Meta + MSFT customer revenue share: see SEC primary + IR links below.

🏭 Supply Chain Coordinates (Part 1-C6 framework)

Upstream (Who ANET Depends On)

Supplier What flows Key dependency
AVGO Tomahawk + Jericho switch ASIC (core silicon) Critical single supplier — ANET relies primarily on AVGO ASIC + in-house EOS software
MRVL Some switch ASIC (secondary) Secondary
TSM Some in-house ASIC + assists AVGO foundry Indirect
COHR + LITE + AAOI Optical modules (plugged into ANET switches) Multi-source; customers often choose
Manufacturing assembly OEMs Switch unit manufacturing Multi-source

Downstream (Who Depends on ANET)

Customer Relationship Concentration
META AI cluster + traditional datacenter network — META is ANET's #1 customer (reportedly ~20%+ of ANET revenue) Highly concentrated
MSFT Azure AI cluster — ANET's #2 customer Concentrated
GOOGL + AMZN + ORCL Hyperscaler network deployments (but GOOGL Jupiter is mostly in-house, AMZN Trainium clusters partially in-house) Critical
Tier-2 clouds + large enterprises + financial institutions Traditional datacenter + HFT Diversified

Competitors

  • AI cluster fabric: NVDA Mellanox InfiniBand + Spectrum-X Ethernet (NVDA's own fabric directly competing with ANET for AI workloads)
  • Traditional switches: CSCO Cisco (traditional enterprise networking leader, but low share in AI datacenter) / HPE Aruba / JNPR Juniper (HPE acquisition completed)
  • White-box switches: Edge-Core / Mellanox-based white-box (hyperscaler in-house SONiC OS trend)
  • Chinese domestic: Huawei (China market)

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

  • Revenue mix: Product (switches + networking hardware) ~85% / Subscription + Service (EOS software + maintenance) ~15% (high-GM recurring revenue)
  • Gross margin: Overall ~64%+ (high for networking; EOS software + service high-GM lift)
  • Pricing power: EOS software + long customer qualification give some leverage — medium-strong
  • Substitution difficulty: Medium — dual-source with NVDA Mellanox at hyperscalers; switching EOS → SONiC takes 1-2 years (META + MSFT have used EOS for years)

(For specific quarterly revenue / customer concentration / capex, see the latest SEC 10-Q.)

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

Type Strength (0-1) Source
Brand 0.6 Recognized leader in datacenter networking, but lower mass-market awareness than CSCO
Switching cost 0.7 EOS software + long customer integration
Network effects 0.4 Not classic network effects, but EOS developer + customer base form an ecosystem
Scale + cost 0.6 AI networking specialist + long-tenured CEO Ullal (not a founder)

Composite: 2.3 / 4 → one of the AI networking leaders, but customer concentration + NVDA fabric competition are core risks

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

  • Trigger 1 (quarterly): Either META or MSFT mega-customer revenue share declines — customer concentration swing
  • Trigger 2 (annual): NVDA Spectrum-X Ethernet significantly displaces ANET in AI clusters — long-term share risk
  • Trigger 3 (low probability): Hyperscaler shifts to white-box SONiC + abandons EOS — mega-customer departure signal
  • Trigger 4 (quarterly): AVGO ASIC supply tightens / price hikes — ANET margin compression
  • Trigger 5 (cyclical): Global datacenter capex downturn — AI cycle swing signal

🔗 Teaching Cross-References

📚 SEC Primary Filings (look it up yourself)

Real-time quote: Yahoo Finance ANET


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