The same neocloud facts read both ways: on September 8 the market picked the “contracts” side, and these two announcements were the story.
How to read this
A quick take on an intraday event; the value is in the information, not the evidence. These are the first-hand press releases behind the day's two catalysts; the note relied on second-hand search of X, and the day's price moves are still second-hand figures to check against a market terminal. The third catalyst, CoreWeave, had no announcement of its own; it rode the sector rotation and the afterglow of earnings.
Palantir will bring Nebius compute and inference inside its security perimeter, so customers can adapt open models with their own data there, and the two will build modular data centers where power exists; no deal value or term was disclosed.
Nebius newsroom: Palantir and Nebius partner to deliver a complete sovereign AI stack to Palantir customers
Palantir names Nebius its preferred sovereign AI infrastructure partner, bringing Nebius compute and inference inside the Palantir perimeter. Strategic partnership based on shared vision that open models and looped customer data create the smartest domain intelligence and provide better security. Companies will work together to bring new AI capacity online faster.
Miami and Amsterdam, September 8, 2026 — Palantir Technologies Inc. (Nasdaq: PLTR) and Nebius Group N.V. (Nasdaq: NBIS) today announced a strategic partnership to bring Nebius’s AI-native compute infrastructure and cloud platform to Palantir’s commercial customers.
As part of the partnership, Palantir has named Nebius its preferred sovereign AI infrastructure partner and, following the integration period, will bring Nebius compute and inference endpoints inside the Palantir enterprise perimeter, enabling eligible customers of Palantir to access Nebius’s cloud and inference infrastructure and thereby giving them control over their compute, data, and models.
The companies will also work together to accelerate the deployment of new compute capacity to serve Palantir customers, including through modular data-center deployments at sites where power is already available.
Palantir believes that AI should serve an organization rather than governing it. Palantir’s Sovereign AI Operating System, built on AIP, Ontology, Foundry and Apollo, provides the authorization and isolation layer that lets organizations train models on their own proprietary data, and retain control of their compute, their models, their data, and the advantage those models produce.
Bringing Nebius infrastructure inside the Palantir perimeter opens that opportunity to commercial organizations that have not previously had a sovereign option to run their own models on trusted infrastructure.
Eligible Palantir customers will be able to deploy open models on Nebius infrastructure and continually adapt them using their own data for their specific domain. This gives organizations the ability to build AI that can outperform general-purpose closed models for their particular use case, while retaining control of both their data and the resulting model.
Alex Karp, co-founder and CEO of Palantir, said: “Nebius’ compute infrastructure powers your ability to run your own AI models under conditions you control. Our ontology and their infrastructure will undergird the sovereignty our partners are demanding.”
Arkady Volozh, founder and CEO of Nebius, said: “Organizations need both the performance of large-scale AI infrastructure and control over their data and models. Together with Palantir, we are bringing this to commercial clients enabling them to run their optimized open models on trusted infrastructure.”
Palantir selected Nebius for what few providers can claim — being built for AI from the ground up rather than adapted from general-purpose computing, with infrastructure and software designed together for demanding AI workloads, and able to be integrated directly into Palantir’s Sovereign AI Operating System.
02
IREN's 2GW hub wins conditional base-load status
Sweetwater 1 and 2, 2GW in total, are conditionally included in ERCOT's Batch Zero; the substation is energized and 300MW is under construction, targeted for delivery in Q4 2027.
GlobeNewswire: IREN’s 2GW Sweetwater Hub Included as Base Load in ERCOT Batch Zero
NEW YORK, Sept. 08, 2026 (GLOBE NEWSWIRE) -- IREN Limited (NASDAQ: IREN) (“IREN”) today announced that its 2GW Sweetwater Hub (Sweetwater 1 and Sweetwater 2) has been conditionally included in the Electric Reliability Council of Texas (“ERCOT”) Batch Zero process as Base Load.
Sweetwater 1 (1,400MW) and Sweetwater 2 (600MW) form part of IREN’s announced >5GW global data center development portfolio.
At Sweetwater 1, IREN’s high-voltage substation was energized earlier this year, and construction of 300MW (gross) of data center capacity continues with delivery targeted for Q4 2027.
Additional large-scale projects within IREN’s broader development pipeline have also been included in Batch Zero. Consistent with its approach to date, IREN will include these projects in its announced development portfolio following the execution of the relevant grid connection agreements.
ERCOT’s classifications remain conditional and subject to ongoing approval processes.
IREN will continue to coordinate closely with relevant transmission and distribution service providers, grid operators, regulators, and local communities as it develops future data center capacity.
IREN is a vertically integrated AI Cloud platform, delivering data centers, compute and software for AI training and inference. IREN’s platform is underpinned by its expansive portfolio of land and grid-connected power in renewable-rich regions across North America, Europe and APAC.
Where Indigo landsFurther
Indigo's conclusion
None of the bear case's triggers lit up that day; what lit up was the bull side of the dial: contracts validated and grid connections landing. The pendulum moved; the clock didn't.
What to remember
The same facts read both ways: on September 8 the market picked contracts, with money moving down from GPU chips to power connections and data-center construction.
IREN moved 2GW from contracted power to conditional base-load status, an energized substation and 300MW under construction: the conversion gauge moved one notch.
The Palantir-Nebius deal endorses the sovereign-AI moat story, but the release gives no value or term; whether it's a big deal or a framework remains to be checked.
None of the bear case's triggers lit up: credit spreads, cash flow, resales, shrinking demand. The pendulum moved; the clock didn't.
Back on the long-running theses
adds to
AI capex as the single engine Market evidence for the side where spending is forced and contracts are real, offsetting the spending panic: a live example of both stories being right.
adds to
The shape of demand: bounded vs unbounded Sovereign AI is specialized demand that can't be outsourced, on the long-horizon, unbounded side; if confirmed, this deal is real demand landing in the neocloud supply layer.
Directly touches several tracked names: CoreWeave (the bear case's core weak node), IREN (evidence of grid conversion), Nebius (the sovereign AI deal), plus Oracle through spillover and Palantir as the partner. An observation, not a rating action.
What would change my mind
credit spreads jump for the AI borrowers, or the backlog stops converting.
Finished. Indigo's take on this piece is in two places: