# data center revenue — X 热门讨论 (2026-09-25 12:23 UTC)

## @wallstengine (Wall St Engine) · 09-25 11:51 · ♥105 ↻12 💬9 $CIFR EXTENDS BARBER LAKE DEAL TO 20YRS

Cipher Digital extended the contracted life of its Barber Lake data center from 10 to 20 years, increasing total contracted revenue from $3.8B to more than $9B.

The structure combines its existing Fluidstack lease with a binding 10-year follow-on commitment from an unnamed AI lab, adding about $5.2B of incremental contracted revenue.

Data halls are expected to come online in phases from Q4 2026 through Q1 2027, with rent starting as each hall is delivered.

Cipher will absorb the first $359.3M of cost overruns above the original budget, while the tenant will reimburse 50% of costs above that threshold through additional rent over the lease term. https://x.com/wallstengine/status/2103452286651641857

## @onechancefreedm (EndGame Macro) · 09-24 22:23 · ♥30 ↻6 💬3 Oracle’s Project Jupiter Force Majeure Notice Exposes the Real Risk Behind Its AI Expansion

Oracle’s Project Jupiter force majeure notice should not be treated as proof that the New Mexico data center campus is being cancelled. But it should not be dismissed as meaningless legal paperwork either.

The notice appears designed to preserve Oracle’s contractual rights if permitting, power or infrastructure delays interfere with performance. Oracle and Blue Owl maintain that financial commitments are unchanged and the project remains on schedule. Those statements matter, but they do not prove that the timing, financing costs or economics remain unchanged.

The Real Vulnerability

Jupiter is enormous, with reported planned capacity around 2 to 2.45 GW. The critical issue is not simply whether Oracle can construct the buildings. It is whether reliable power reaches them when the customer needs the capacity.

The project relies heavily on Bloom Energy fuel cells and natural gas infrastructure. Pipeline permitting and New Mexico land issues have already complicated part of the power plan.

A building can be completed before the infrastructure necessary to monetize it is operational.

That distinction matters because Oracle can commit capital, leases, equipment and financing long before it receives the cloud revenue expected to justify those costs.

The Financing Mismatch

Oracle has roughly $664 billion of remaining performance obligations, yet only about 13% is expected to convert into revenue within 12 months.

It also has approximately $288 billion of additional uncommenced lease commitments, generally extending 15 to 19 years.

Those figures measure different things, but together they expose the fundamental timing problem.

Oracle is taking on very long obligations while much of the revenue expected to support them remains years away.

The latest quarter makes the cash intensity clearer.

Operating cash flow was roughly $23 billion.

Cash capital spending was about $28.5 billion.

Conventional free cash flow was therefore negative by roughly $5.4 billion.

Operating cash flow also included approximately $11.4 billion of customer prepayments with a financing component. That cash is valuable, but it arrives before Oracle has delivered all the corresponding services.

AI bookings and economically earned cash are not the same thing.

Why The Cycle Matters

This becomes more dangerous in a late cycle environment.

If financing costs remain high, lenders demand greater protection and customers become more selective, Oracle can find itself financing infrastructure today for revenue expected years later.

A 12 to 24 month delay does not have to kill Jupiter to damage the economics.

Revenue arrives later.

Financing costs continue.

Equipment ages.

Customer bargaining power increases.

Construction costs accumulate.

The market applies a larger discount to future AI revenue.

Oracle does not need AI demand to collapse for shareholders to suffer.

AI demand can remain enormous while Oracle earns disappointing returns on the infrastructure required to serve it.

My Take

Jupiter exposes Oracle’s dependence on infrastructure it does not completely control.

Power approvals, pipelines, contractors, lenders and major customers all sit between Oracle’s commitments and the eventual cash flow.

The real question is therefore not whether AI will grow.

It is whether Oracle can convert massive contracted demand into profitable cash flow quickly enough to justify the capital, leases and financing committed before that revenue arrives.

Project Jupiter does not prove Oracle is failing. It exposes how little room Oracle has for delay when so much of its AI strategy depends on capital being deployed years before the final economics are known. https://x.com/onechancefreedm/status/2103248950530568203