Oracle (ORCL)

AI cloud infrastructure (OCI) + database + enterprise apps — the database franchise now funding an AI hyperscaler

AI exposure 31% — derived from reported segment revenue, not asserted.

Deep-dive research

AI Infrastructure Software: Who Gets Paid When Enterprise AI Runs in Production? — an industry deep dive with its slide deck, published on Frontier Macro & Markets, 2026-09-22, covering SNOW, MDB, ORCL, NET, DDOG, DT.

Where it sits

AI Platforms → Hyperscalers & Cloud Platforms · figures rebased to Q1 FY2027, reported 2026-09-10

Cloud infrastructure (OCI) is 38.2% of revenue after the Q1 FY2027 rebase and the largest single business Oracle has, so placement rule 2 — the primary follows the largest business — puts Oracle with the hyperscalers. The rest is tagged separately: Enterprise & Business Software at 29.1% (cloud applications 21.8% plus services 7.3%); Data & Infrastructure Software at 28.7%; Storage & Server Systems at 4.0%.

Also tagged: secondary in Software Stack → Enterprise & Business Software; secondary in Software Stack → Data & Infrastructure Software; emerging in Physical Infrastructure → Storage & Server Systems.

Reported segments

Oracle ($19,345M Q1 FY2027, +30% YoY). Rebuilt on the five revenue lines Oracle reports, replacing a model that carried the cloud infrastructure / cloud applications split as an estimate — and had it backwards, with applications larger, when Oracle's own Q4 figures already showed infrastructure ahead. Cloud infrastructure (OCI) 38.2% ($7,388M, +121%) = primary B6 Hyperscalers & Cloud Platforms, and the largest business in the company; software licence and support 28.7% ($5,550M, -3%); cloud applications 21.8% ($4,219M, +10%); services 7.3% ($1,414M); hardware 4.0% ($774M). Oracle's ASC 280…

How the 31% is built

Investment read

Moat. Database lock-in that funds an AI hyperscaler: software licence and support is still $5.550B a quarter, and cloud infrastructure has become the largest business at $7.388B, up 121%, on a $664B order book. The lock-in is real in the database; the infrastructure half is a scale-and-capacity business competing with Microsoft, Amazon, Google and the neoclouds for the same AI contracts.

Bottleneck / pricing power. Two pricing regimes pulling apart. Database lock-in is structural pricing power; AI capacity is priced to win share, and the filing shows the effect — the cloud & software segment margin fell to 54.5% from 59.6% as infrastructure grew 121%. Management says demand grows faster than supply, which supports price, but the margin line is moving down, not up.

Role in the AI stack. The database-funded AI hyperscaler — GPU capacity for the AI labs, sold alongside the enterprise database and applications estate.

Connected companies

Others in Hyperscalers & Cloud Platforms

All 176 companies · What moved this quarter · How the map is built