AI's $600B Question

Notes

Sequoia partner analysis; reasoned napkin-math argument rather than primary data. Slightly above blog prior (0.35) because author is a VC with direct visibility into AI infrastructure spend and the core inputs (Nvidia run-rate, hyperscaler capex, OpenAI revenue) are checkable against third-party reporting cited in-piece.

Evidence from this source (5)

• CDAs of mid-2024, applying 2x to Nvidia's AI data center run-rate revenue (to acc… support
  weight

  0.70

  method: expert_estimate · locator: opening section; 'AI's $200B question is now AI's $600B question'

  “All you have to do is to take Nvidia's run-rate revenue forecast and multiply it by 2x to reflect the total cost of AI data centers... Then you multiply by 2x again, to reflect a 50% gross margin for the end-user of the GPU.”

• CDMicrosoft alone accounted for approximately 22% of Nvidia's Q4 FY2024 revenue,… support
  weight

  0.65

  method: triangulation · locator: section 'GPU stockpiles are growing'

  “Microsoft alone likely represented approximately 22% of Nvidia's Q4 revenue.”

• CDThe acute GPU supply shortage that peaked in late 2023 had largely subsided by… support
  weight

  0.55

  method: primary_testimony · locator: section 'The supply shortage has subsided'

  “For most people I speak with, it's relatively easy to get GPUs now with reasonable lead times.”

• CDOpenAI's annualized revenue reached approximately $3.4B by mid-2024, up from $1… support
  weight

  0.70

  method: journalistic_report · locator: section 'OpenAI still has the lion's share of AI revenue'

  “OpenAI's revenue is now $3.4B, up from $1.6B in late 2023.”

• CDGPU cloud compute is commoditizing toward per-hour metered pricing with low pri… support
  weight

  0.50

  method: expert_estimate · locator: section 'Lack of pricing power'

  “GPU computing is increasingly turning into a commodity, metered per hour. Unlike the CPU cloud, which became an oligopoly, new entrants building dedicated AI clouds continue to flood the market.”