US corporate AI spend shows 600x gap as Neoclouds ride crypto infrastructure into AI

Ramp credit card data shows US corporate AI spending is accelerating sharply and highly bifurcated: in July, the top 1% of companies by AI spend invested a record median of roughly $7,400–$7,500 per employee per month, the top 10% about $650–$660, while the median company only about $12—a gap of more than 600x. Around the same time, a16z published charts and analysis arguing that cloud computing is splitting into general-purpose clouds and Neoclouds (emerging specialized AI clouds), with former cryptocurrency mining companies leveraging existing power rights and infrastructure expertise to pivot into the AI compute market.

Confirmed

• Ramp data (consistent across multiple posters): top 1% of companies spent roughly $7,400–$7,500 per employee per month on AI, the top 10% about $650–$660, and the median company about $11.95–$12
• @rohanpaulai calculated the gap at roughly 625x—top companies can spend in one afternoon what the median company budgets for AI in an entire year
• a16z's cloud industry charts and article divide the cloud into general-purpose clouds and Neoclouds, analyzing the rise of Neoclouds

Why it matters

• @rohanpaulai argues that even if median companies move a small step closer to the top, inference demand would increase substantially, making this spending distribution itself an important forward-looking signal of inference demand growth
• @DavidLinthicum noted in a quote-tweet analysis that the core requirement of AI training is not GPU chips per se, but making thousands of accelerators compute in sync like a single machine—imposing extreme demands on compute continuity, high-speed interconnects, storage, scheduling, power, and cooling—so what's scarce is these systems engineering capabilities, not the chips themselves
• a16z's article draws historical analogies (railroad companies laying fiber along their rights-of-way to become Sprint, natural gas pipeline companies laying fiber to become WorldCom), arguing that crypto miners with power rights, infrastructure, and construction experience have a natural advantage in transforming into Neoclouds—a industrial migration path worth watching

2026-08-17 ~ 2026-08-18 · 5 related posts

• [source] AI Spending Gap 625x: Top 1% Spend $7,500/Employee/Month vs Median $12 — rohanpaul_ai · 2026-08-17
• Top 1% US firms spend record $7,400 per employee on AI as investment concentrates — luisdans · 2026-08-17
• [source] a16z Charts: How Neoclouds Repurpose Crypto Infrastructure for AI — FinanceYF5 · 2026-08-18
• Neoclouds Rise as AI Demands Synchronized Compute, Not Just Chips — DavidLinthicum · 2026-08-18

1 near-duplicate retellings: FinanceYF5