AI Infrastructure Data
These are the charts behind the FinSight blog: AI compute supply, cloud earnings, capital spending, financing, and the physical bottlenecks in between. Every chart has a fixed address, a one-line takeaway and its source, and the links stay put when the data updates. The articles carry the views; this site carries the numbers.
Three lines: demand, supply and spending, financing
Demand
Is demand real? Hypothesis → earnings calls → prices → revenue
- Demand hypothesis: compute is tightSupply and demand are tightest around mid-2027; the compute cost curve crosses the cost band a year out. Supply-Demand Scenarios and the Cost Line →
- Earnings check: what the calls said2Q26 earnings calls: the shortage runs to end-2027 and gets more valuable. Three Public Clouds: Revenue and YoY Growth →
- Price check: pricing tiersPrice cuts are not a retreat, they are tiering: the frontier costs more, old models approach free. Flagship Model Price History →
- Revenue: ARR and productsGrowth is a step-function: a new product opens a new customer group, and that is the next step. ARR Path and Product Launches →
- Enterprise adoption: AI inside companiesBreadth shows in adoption rates and job postings; depth shows in who gets the enterprise dollars. Adoption Rates at a Glance →
Supply and Spending
Is the money turning into compute? Spent → macro → shipments → built
- Spending: CapEx structureReported CapEx understates the real spend more and more; use all-in investment (CapEx + new leases). Quarterly CapEx and All-In Investment →
- Supply baseline: Epoch trackingHow many chips shipped, how much compute has accumulated, and whether prices and monetization line up. Performance vs Price Frontier →
- Macro cross-check: equipment investment itselfIt shows up in the macro accounts: US equipment investment $798B annualized, up 31.6%, consistent with the four hyperscalers's $726B all-in investment. Macro Cross-Check, Layer by Layer →
- Supply chain check: Taiwan, Korea, USThe money did turn into shipments, confirmed at four levels. Coming soon
- Physical side: power and equipment bottlenecksIt is not that they don't want to build, equipment can't keep up: an execution bottleneck, not weakening demand. Coming soon
Financing
Can the money keep up? Where it comes from → who it concentrates in → where we are in the cycle
- Financing: the chain and credit tiersNext year's refinancing is small: on a debt-only basis the runway reaches end-2028 at about 1.07x. CSP Cash Flow and Headroom →
- Concentration: off-balance-sheet, circular deals, NVDA backstopsQuantity is fine. Pricier tiers, deals moving off balance sheet and NVDA backstops add up to concentrated risk. Circular Deals Map →
- The 30-year anchor and who's buying the bondsThe 30-year at new highs is money moving, not money running out: AI bond deals are still oversubscribed. Coming soon
- Historical frame: where we are in the cycleA five-year investment cycle, and cycles die on the financing side. Coming soon
Recently updated chart pages
- Supply and Demand: Four Lines, One Frame2026-09-04
- ARR Path and Product Launches2026-09-04
- ARR Growth per Month2026-09-04
- New ARR per New Chip2026-09-04
- Breadth and Depth: How Wide, How Deep2026-09-04
- Adoption Rates at a Glance2026-09-04
- Model Vendors: Who's Getting the Enterprise Dollars2026-09-04
- AI Spend per Employee2026-09-04
- App Categories: Penetration and Retention2026-09-04
- OpenAI Signals: Job Functions and Consumer2026-09-04