SB Energy IPO Filing Reveals 8.8 GW of Contracted AI Data Center Capacity
SB Energy has publicly filed for a U.S. IPO, revealing an 8.8 GW contracted data center portfolio and a strategy that combines gigawatt-scale computing campuses with dedicated power infrastructure for customers including OpenAI and SoftBank.
Microgrid Media Editorial Desk
Published · Updated · 5 min read
The SB Energy IPO filing reveals an 8.8 GW contracted AI data center portfolio spanning Ohio and Texas, much of it tied to OpenAI. The September 1 filing also shows how the SoftBank-backed company plans to combine hyperscale computing campuses with dedicated generation, storage and transmission infrastructure.
The Form S-1 filed with the Securities and Exchange Commission shows that SB Energy is positioning itself as an integrated developer of both data centers and power generation rather than a conventional renewable-energy company alone. Reuters reported that the company has approximately 8.8 GW of contracted data center capacity while much of that portfolio remains under development rather than operating today.
The company is building around a power-first data center model
SB Energy describes its strategy as integrating data center development, construction and operation with power-generation assets capable of supporting large computing loads.
The company’s portfolio historically centered on renewable energy, including solar and battery storage, but its current growth strategy increasingly revolves around AI infrastructure requiring multi-gigawatt electricity supply.
This broader shift is examined in Microgrid Media’s guide to data center microgrids, which explains how on-site generation, battery storage and advanced controls can be developed alongside utility service.
That approach reflects a broader change in the data center market. Developers increasingly are trying to secure land, generation, transmission access and computing infrastructure together rather than waiting for conventional utility interconnection processes to deliver all required capacity.
Microgrid Media has been tracking the same trend through projects including OpenAI’s 3.2 GW Georgia Power agreement, which combines one of the largest planned AI loads in the Southeast with utility infrastructure and up to 1 GW of flexible demand response.
PORTS-Pike could become one of the world’s largest AI campuses
A major portion of SB Energy’s future portfolio centers on the PORTS-Pike Technology Campus in Pike County, Ohio.
SB Energy said in an August 17 project announcement that NVIDIA will provide the AI compute infrastructure for the site while OpenAI will be the customer for 8 IT GW of capacity.
SB Energy plans to build, own and operate the data center under a 20-year lease with OpenAI. NVIDIA initially secured 4.25 IT GW of land, power and shell capacity and holds an option associated with another 3.75 IT GW, according to the company.
The Ohio development demonstrates how extreme the scale of emerging AI electricity demand has become. Individual campuses are increasingly being discussed in gigawatts rather than the tens or hundreds of megawatts historically associated with major data centers.
Microgrid Media has also examined how long-duration energy storage is being incorporated into data center power strategies, including Google’s planned use of 100-hour batteries alongside renewable generation.
OpenAI is also tied to a 1.2 GW Texas campus
The Ohio development is not SB Energy’s only OpenAI project.
In January, OpenAI and SoftBank each agreed to invest $500 million into SB Energy as part of a broader Stargate infrastructure relationship. OpenAI also selected the company to build and operate a 1.2 GW data center site in Milam County, Texas.
SB Energy’s January partnership announcement said the company was developing multiple multi-gigawatt campuses and expected its first facilities to begin entering service during 2026.
Those projects create a direct relationship between computing capacity and power-development capability. Rather than treating energy procurement as a secondary operating expense, SB Energy is making electricity infrastructure part of the core data center development model.
The IPO filing exposes the scale and financial risk
The proposed offering also gives investors a clearer look at the financial commitments required to build this infrastructure.
Reuters reported on September 1 that SB Energy generated approximately $138.7 million of revenue in the first half of 2026 while recording a net loss of about $3.21 billion.
The company’s future economics therefore depend heavily on projects that are contracted or under construction but not yet fully operational.
The IPO itself also remains proposed. SB Energy has applied to list under the ticker SBE on Nasdaq and Nasdaq Texas, but the number of shares and offering price had not been established in the public filing reviewed for this article.
Power availability is becoming the limiting factor for AI growth
The scale disclosed in the filing reinforces a central issue facing utilities and grid planners: the pace at which AI companies want computing capacity can exceed the pace at which conventional generation, transmission and distribution infrastructure can be built.
This creates incentives for data center developers to secure dedicated generation, storage and flexible-load arrangements alongside their campuses.
The SB Energy data center power model goes further by placing the development of computing infrastructure and power infrastructure under the same platform.
That may reduce some coordination delays, but it does not eliminate interconnection, permitting, fuel, transmission or reliability constraints. Projects measured in multiple gigawatts still require enormous physical energy systems, even when generation and data center development are planned together.
What happens next
The IPO process will provide additional financial disclosures as the registration statement advances and underwriters establish an offering size and price range.
For the power sector, the more important milestones will be physical rather than financial: how quickly SB Energy can energize its contracted campuses, what generation mix supports those loads, and how much new transmission or on-site power infrastructure is required.
The Ohio and Texas campuses will also provide major tests of whether vertically integrated data center developers can deliver multi-gigawatt computing capacity faster than traditional utility-led expansion while still maintaining grid reliability and protecting existing customers from infrastructure costs created by new AI loads.
The Microgrid Media Editorial Desk reports on microgrids, energy storage, grid modernization, data center power demand, distributed energy resources, and energy policy. Reporting published under the Editorial Desk byline is developed from primary regulatory records, utility filings, government documents, company disclosures, and independently verifiable sources, and is reviewed under Microgrid Media’s editorial standards.
Editorial oversight: Jonas Muthoni, Editor-in-Chief
