Oracle just reported a quarter that is difficult to fit into any conventional cloud-company narrative: $19.35 billion in revenue, a 121% surge in infrastructure revenue, and more than $30 billion in new AI cloud contracts signed in 90 days, pushing total contracted future revenue to $664 billion.
The results, reported on 10 September 2026, arrived at a moment when markets are actively debating whether the enterprise AI buildout represents durable, multi-year infrastructure investment or a front-loaded spending cycle that could unwind. Oracle’s numbers inject fresh evidence into that debate, but the picture is not clean: the company also posted negative free cash flow for the sixth consecutive quarter and carries a capital expenditure plan of up to $95 billion for the full fiscal year.
The sections below break down the headline figures, the guidance revision, the backlog mechanics, and the specific risk factors worth weighing before treating that $664 billion figure as a straightforward indicator of future earnings power.
Oracle’s Q1 results: what the numbers actually show
Total revenue landed at $19.35 billion, up 30% year-over-year and ahead of the $19.13 billion Wall Street consensus. Adjusted earnings came in at $1.92 per share, a 30% annual increase that beat the $1.73 consensus by roughly 11%. On a GAAP basis, diluted EPS reached $1.56, up 55%.
The market’s immediate verdict was positive: shares advanced approximately 6.9% in after-hours trading. Profitability moved in the right direction too, with non-GAAP operating income of $8.2 billion (up 31%) and GAAP operating income of roughly $6.7 billion (up 57%). This was not a revenue-line beat propped up by shrinking margins; the earnings improvement was real.
The pre-earnings expectations heading into this report included Wall Street consensus of $1.74 adjusted EPS and $19.13 billion in revenue, alongside options markets implying an 11-11.7% post-earnings swing, a range that the actual 6.9% after-hours advance came in beneath despite the headline beat.
| Metric | Q1 FY2027 | YoY Change |
|---|---|---|
| Total revenue | $19.35B | +30% |
| Non-GAAP EPS | $1.92 | +30% |
| GAAP EPS | $1.56 | +55% |
| Total cloud revenue | $11.6B | +62% |
| OCI (IaaS) revenue | $7.4B | +121% |
| SaaS revenue | $4.2B | +10% |
| Software (non-cloud) revenue | ~$5.5B | -3% |
| GAAP operating income | ~$6.7B | +57% |
Cloud infrastructure versus legacy software: a company in two speeds
Look beneath the headline and the growth is anything but uniform. Oracle Cloud Infrastructure (OCI), the company’s rented computing and storage business, grew 121% to $7.4 billion. Its software-as-a-service applications business, by contrast, grew a comparatively modest 10% to $4.2 billion.
Then there is the part the bullish read tends to skip. Software non-cloud revenue, Oracle’s traditional licensing base, fell approximately 3% to around $5.5 billion. The legacy franchise is not merely growing slowly; it is contracting.
That split matters for how you weigh the stock. A 121% infrastructure sprint alongside a shrinking legacy base tells you the AI bet is now load-bearing, carrying the story on its own. It is a concentration risk worth tracking every quarter from here.
When big ASX news breaks, our subscribers know first
A $664 billion backlog, and the question of when it pays out
The number that stopped everyone was the backlog. Remaining performance obligations (RPO), which is contracted future revenue Oracle has signed but not yet booked, hit a record $664 billion. A year earlier that figure sat at $455 billion; a single quarter earlier, in Q4 FY2026, it was $638 billion.
Then the mechanics arrive, and they reframe the headline. Chief Financial Officer Hilary Maxson told investors that only about half of the $664 billion is expected to convert into recognised revenue over the next 36 months. The rest, including the most recently signed AI agreements, is not scheduled to affect revenue or capital requirements until FY2028 or later.
Here is how the RPO actually breaks down:
- What it is: signed, contractually committed customer revenue, not a sales pipeline of prospects.
- The near-term slice: roughly 50% is expected to convert within 36 months.
- The long tail: the most recently booked AI contracts are structured to leave financials untouched until FY2028 or later.
Maxson’s framing is the single most important qualifier on the whole figure.
Approximately 50% of the $664 billion in remaining performance obligations is expected to convert into revenue over the next 36 months, with the most recently signed AI cloud contracts not impacting revenue or capital requirements until FY2028 or later, according to CFO Hilary Maxson.
The scale is genuinely real, and Oracle is already delivering against it: the company added around 850 megawatts of new data centre capacity and more than 300,000 GPUs in the quarter alone. But if you are reading $664 billion as near-term revenue visibility, adjust. A large portion of it is long-dated, which means the current share price is pricing in execution measured in years, not quarters. That timing detail is the variable most market commentary underweights.
Counterparty concentration risk sits beneath the backlog figure in a way the headline number does not surface: a single customer relationship representing a substantial fraction of contracted obligations means the durability of that $664 billion depends as much on one counterparty’s revenue trajectory as on Oracle’s own execution.
Guidance raised, but cash flow tells a different story
Management backed the quarter with an upgraded outlook. Oracle now guides to at least $90 billion in full-year FY2027 revenue, roughly 34% growth in constant currency, and lifted its non-GAAP EPS target to $8.10 from $8.05. For the coming quarter, it guided to revenue of $20.9 billion to $21.5 billion and adjusted EPS of $1.85 to $1.93, broadly in line with the roughly $21.2 billion and $1.89 Wall Street had modelled.
| Guidance metric | New guidance | Prior / consensus |
|---|---|---|
| FY2027 revenue | ≥$90B (~34% cc growth) | Raised |
| FY2027 non-GAAP EPS | $8.10 | $8.05 (prior) |
| Q2 FY2027 revenue | $20.9B-$21.5B | ~$21.2B consensus |
| Q2 FY2027 non-GAAP EPS | $1.85-$1.93 | $1.89 consensus |
A guidance raise is a confidence signal. The problem is that it sits directly on top of a cash flow picture that complicates the bullish case.
Six consecutive quarters of negative free cash flow: context and concern
Free cash flow in Q1 was negative $5 billion. That is the sixth straight quarter Oracle has burned more cash than it generated after capital spending. Capital expenditure ran to roughly $28 billion to $28.5 billion in the quarter alone, against a full-year plan of up to $95 billion.
The tension is easiest to see in one comparison. Operating cash flow hit a record $23 billion, up 184% year-over-year, yet free cash flow was still negative $5 billion. Operating cash flow measures the cash the business throws off before capital investment; free cash flow is what remains after it. Oracle is generating enormous operating cash and then ploughing all of it, plus more, back into infrastructure.
The tension between free cash flow versus EPS is structural for capital-intensive businesses in a build cycle: operating cash flow measures what the business earns before it reinvests, while free cash flow measures what remains after, and Oracle’s record $23 billion operating cash flow alongside negative $5 billion free cash flow is a textbook illustration of why the two figures require separate treatment.
That is either an exceptionally well-timed bet or a heavy concentration of capital risk, depending entirely on how AI demand evolves from here. Oracle has stated it does not expect newly structured contracts to require additional capital raises beyond existing plans, but a near-$100 billion annual capex programme warrants close scrutiny regardless. For anyone valuing the stock today, the guidance raise and six straight negative quarters together define the risk-reward on offer.
What Oracle’s quarter signals about enterprise AI infrastructure demand
Step back from Oracle specifically and the quarter reads as a datapoint about the entire sector. Oracle sits alongside Nvidia, Microsoft, Google and Meta as a publicly traded gauge of how deep the enterprise AI buildout runs, and these numbers argue it is deepening, not fading.
The clearest tell is in how the deals are structured. Many of Oracle’s largest AI contracts use bring-your-own-hardware or prepaid arrangements, where customers install their own GPUs in Oracle facilities and pay for power, cooling, networking and platform services. Customers accepting terms like that are behaving like buyers in a shortage, not a surplus.
Three structural demand signals show up in the quarter:
- Contract scale and structure: more than $30 billion in new AI contracts booked in 90 days, many on non-standard prepaid or bring-your-own-hardware terms.
- Supply constraint: Oracle states demand for AI training and inference capacity is expanding faster than available supply, pushing customers to lock in capacity years early.
- Management confidence: the raised FY2027 revenue guidance of at least $90 billion, backed by delivery of 850 megawatts and 300,000-plus GPUs in a single quarter.
Customer demand for AI cloud training and inference services is expanding at a pace that exceeds available supply, with customers accepting non-traditional deal structures to secure data centre capacity, according to Oracle management commentary accompanying the results.
For anyone tracking the buildout broadly, the read is that long-term, non-cancellable compute commitments are becoming standard operating procedure among AI firms. That has implications for every infrastructure provider in the sector, not just Oracle.
The caution is that the theme and the stock are not the same trade. Oracle’s contracting legacy software, its balance sheet leverage, and its execution risk across a multi-year buildout are company-specific factors. A bullish view on AI infrastructure as a sector does not automatically translate into a bullish view on Oracle at its current valuation. The two require separate analytical frameworks.
What the results change, and what they leave unresolved for Oracle investors
The quarter gives Oracle bulls firmer footing than they had 90 days ago. It does not settle the argument. The central open question is whether a capital commitment approaching $95 billion a year will earn returns that justify the risk before AI demand or pricing shifts underneath it.
The market’s immediate answer was a 6.9% after-hours advance. A single session, however, cannot resolve a thesis that hinges on execution stretching into FY2028, when the most recently signed contracts first begin flowing through the financials.
Three data series will tell you far more than the next headline beat:
- Free cash flow trajectory: whether the six-quarter negative streak narrows or extends.
- RPO conversion rate: how much of the $664 billion starts converting to recognised revenue on the 50%-in-36-months schedule.
- Legacy software revenue: whether the 3% decline stabilises or accelerates.
Treating the beat and the guidance raise as reason enough to act, without working through the backlog timeline and cash flow path, means pricing in execution risk you may not fully see. The next two to three quarters are the real test of whether $664 billion is a durable earnings engine or a deferred question.
For investors wanting to model whether the current buildout creates or destroys shareholder value, our full explainer on Oracle’s capital intensity metrics covers the five key ratios to track, including free cash flow trajectory, gross margin progression, and the equity issuance pace that introduces dilution even when revenue grows.
This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions.
Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors, and forward-looking statements are subject to change based on market developments and company performance.
