Whale Rock Capital Management · S-curve / technology-adoption tech investor (ex-Fidelity) — running synthesis of his appearances, with per-transcript breakdowns and a stock index.
Highest-conviction position (bought Aug 2025 at ~$180B) — coding is the "true unlock" (~$0.5T market), with critical IP, an enterprise brand and recursive self-improvement.
Decommoditized critical infrastructure — sole Google-TPU-server supplier, ~50-60% of cloud Ethernet switches, liquid-cooling lead; bought ~3 yrs ago at ~8x.
Decommoditized printed circuit boards — AI servers need more boards at far higher layer counts (10→120), lifting units + ASPs; supplies Google/Nvidia ("just won Nvidia") + ~40% defense (Iron Dome).
Emblem of software under pressure — AI only ~1-2% of ~$40B sales, with budget/seat headwinds and "headless"/database-relegation risk; Whale Rock net short software.
In one line: Whale Rock buys where a big S-curve (technology-adoption curve), a durable competitive advantage, and underappreciated long-term earnings power all line up — letting it own the best companies cheaply (Nvidia ~4×, Tesla ~5×, Amazon "for free"). AI is the biggest S-curve ever; own the chips/infrastructure and the model-layer oligopoly, avoid application software.
The framework — three things must line up. A big S-curve gives exponential unit growth; a strong moat protects it; and the market under-appreciates the long-term earnings power (earnings going $1 → $10 → $20). Because the world thinks linearly and short-term, this repeatedly lets Whale Rock buy great companies at low P/Es and underwrite 2-3 years out.
AI is the biggest S-curve ever — chips → models → apps. The new compute stack runs power → chips → clouds → foundational models → applications. Whale Rock went into chips/infrastructure first (they get demand first and you know the winners); the infrastructure layer is only ~10% penetrated and the enterprise application layer <1% — "an L-curve, just straight up." There isn't enough compute in the world.
The model layer is a three-horse oligopoly. ~60 contenders winnowed to three — Anthropic (his highest-conviction position; bought Aug 2025 at ~$180B), Google (Gemini), and OpenAI. Coding is the "true unlock" (~$0.5T market from coding alone). Models aren't commodities: critical IP, an enterprise brand, escape-velocity scale, the "harness" ecosystem, and recursive self-improvement defend the leaders.
Decommoditization of AI hardware — own the pinch points. AI workloads grow ~10× a year and push every part of the rack to its limit, turning once-commodity suppliers into IP-rich, high-margin choke points: Celestica (TPU servers, Ethernet switches, liquid cooling), Corning (fiber), power supplies (Advanced Energy/Delta), PCB/Elite Material, memory/HBM, and scale-up/across/over-fiber. "One of the best ways to play AI."
Bearish application software. Sold almost all of it and entered the year net short. AI products aren't moving the needle; faster-ROI token spend squeezes software budgets; pricing power and seat-based models are under threat; and "headless"/build-it-yourself risk looms (Salesforce the emblem).
Large-cap-tech alpha — the Mega-Cap Tech Fund. A structural underweight of the world's largest tech companies; much of Whale Rock's historic performance came from the biggest names (Apple, Amazon, Tesla). The new fund's universe is the top-30 global market caps, picking the best ~12-13 wide-moat winners (Nvidia, TSM, SK Hynix, ASML).
Transcripts
One dated page per appearance — each has its full stock table, talking points, and the saved transcript. Newest first.