a16z unpacks 25 charts: hyperscaler CapEx nears $1T annually as demand outruns supply
a16z · youtube · 2026-09-30
a16z investors David George, Sarah Wang, Alex Immerman and Santiago Rodriguez walk through 25 charts from their latest State of Markets presentation on AI infrastructure and the next investment cycle.
Infrastructure & macro
- Hyperscaler CapEx is approaching $1 trillion annually ($780B race), with compute demand still outrunning supply
- Rising markets have so far been supported by earnings rather than multiple expansion
- Spending ripples through chips, power, construction and physical infrastructure
Applications & enterprise
- OpenAI and Anthropic revenue growth; only 2% of enterprise AI deployment is truly tracked
- Power users spend 20x the median; the rise of agents and falling inference costs
- Amazon blocking Muse, Instacart opening up, the SaaS bifurcation, and Stripe's Renaissance data
What's next: consumer agents, robotics, autonomy, AI × biology, personal health, defense, and continued enterprise AI diffusion. Full timestamps included.
Related event: a16z: Hyperscaler Capex Nears $780B as AI Cycle Unfolds(4 posts)→
More from Venture
- OpenAI bought TBPN, a16z has 250K subs: tech firms pour millions into owned media — danshipper · 2026-10-01
- How the OpenAI marketplace differs from hyperscaler marketplaces and co-sell motions — matt_slotnick · 2026-10-01
- Kled V3 deploys custom AI data collection to 500K+ contributors in 72 hours — testingcatalog · 2026-09-30
- ElevenLabs employee tender values the company at $22B, 2x since February — austin_malerba · 2026-09-30
- Delphi Launches Predict Anything: Forecast Any Claim, Open an On-Chain Market — benfielding · 2026-09-30
- Making the case for AI capex: fast ROI, general utility, and self-replicating capital — menhguin · 2026-09-30