AI Cloud Infrastructure Begins Competing for “Power Financing”
On July 16, Industrial Development Funding and Oaktree announced a $1.7 billion project investment for Nebius’s AI cloud infrastructure. The capital is not going into GPUs themselves, but into dedicated power systems that can be deployed faster.
What happened
On July 16, Industrial Development Funding (IDF) and alternative asset manager Oaktree announced that they will invest $1.7 billion in Nebius’s AI cloud infrastructure project to deploy Bloom Energy fuel cell systems. Once completed, the project will provide Nebius with “behind-the-meter power” — electricity generated and used directly on the data center customer side, rather than waiting entirely for public grid expansion.
The structure of the deal is also notable: IDF is serving as the lead project developer, with Oaktree participating as a minority equity investor; Morgan Stanley is handling tax equity financing and placement, and MUFG Bank is providing senior debt financing. In other words, AI infrastructure financing is expanding from servers, chips, and data center buildings to a set of power assets that can be financed, built, and operated independently.
Nebius had previously announced a partnership with Bloom Energy to adopt fuel cell technology in its U.S. AI infrastructure expansion. The company’s core rationale is deployment speed, power supply capacity, and support for the performance and availability of AI workloads. This $1.7 billion project investment now takes that technology partnership one step further at the capital-structure level.
Why it matters
The bottleneck for AI cloud services is no longer just “whether there are enough GPUs.” Model training and inference require high-density, continuous, and predictable power, while large data center connections to the grid are often constrained by interconnection queues, transmission buildout, and permitting processes. The behind-the-meter power mechanism places part of the electricity generation capacity directly next to the load, reducing dependence on new transmission capacity and the pace of public-grid delivery.
This is also a transaction with market-structure significance. On July 15, Reuters reported that Wall Street banks were seeing rising equity, debt, and project-finance demand driven by AI infrastructure. One day later, the Nebius project involving IDF, Oaktree, Morgan Stanley, and MUFG showed another funding path: infrastructure capital holds the power assets, construction is completed through tax equity and bank debt, and the result serves AI cloud computing demand.
For investors, the lens for observing AI capital expenditures is therefore widening. In the past, the market often measured sector momentum by GPU purchases, cloud providers’ capital expenditures, or data center valuations; now, fuel cells, nuclear power, renewable energy, transmission, and project-finance conditions may also become key variables determining whether compute capacity can come online on schedule. The willingness of capital to price “faster access to power” on a standalone basis shows that power delivery time is becoming a scarce asset in AI infrastructure.
What still needs to be watched
First, the specific construction progress, commissioning timeline, and actual power-supply scale corresponding to the $1.7 billion still need ongoing tracking. The press release confirmed the investment arrangement and financing roles, but did not disclose the full project return, debt tenor, fuel costs, or Nebius’s long-term power purchase terms.
Second, whether fuel cells can reliably meet the around-the-clock load of AI data centers will need to be verified through actual operations. Power delivery speed is only the starting point; system uptime, maintenance cycles, fuel sourcing, and unit power costs will ultimately be reflected in the gross margin of AI cloud services.
Third, whether the project will expand from a single deployment into a replicable financing template. If more behind-the-meter power projects backed by institutional capital emerge later, AI infrastructure could form a new combination of “compute leasing + power assets + structured financing”; if the project is delayed or its economics fall short of expectations, the capital market will also reassess leverage risk in the AI buildout cycle.
Sources
- Industrial Development Funding and Oaktree announce $1.7 billion project investment in Bloom Energy fuel cells for Nebius AI infrastructure build-out (primary source)
- Nebius and Bloom Energy announce partnership to power AI infrastructure build-out
- Reuters: Wall Street banks see AI “super cycle” set to boost deals, financing
Information only. Not investment, legal, tax, or financial advice.