How Nvidia’s SB Energy Deal Unlocks OpenAI’s Ohio Campus

Nvidia SB Energy OpenAI data center: NVIDIA is investing $1.5 billion of equity in SB Energy while providing large-scale credit and residual-value guarantees to back OpenAI’s lease commitments, securing roughly 4.25 gigawatts of IT-load LPS capacity initially and phase-in capacity from 2028 under a 20-year tenant lease to OpenAI [S2][S1].

Why This Matters

The deal is unusual because it separates three different financial roles: equity, lease guarantees, and power/land commitments. That split changes who carries which risk and who controls which piece of the physical build - and it matters for careers because different teams hire into those exact gaps: project finance, lease operations, power procurement, and site buildout management.

Credit: Photo by valaymtw on Pixabay

Nvidia SB Energy OpenAI data center financing: equity vs guarantees

NVIDIA’s $1.5 billion is an equity-style investment in SB Energy; the separate credit support and residual-value guarantees are conditional commitments that back leases rather than being direct construction cash. That distinction is the heart of the structure. Put simply: equity buys a stake and alignment; guarantees lower the financing bar for SB Energy and the lessor by promising a minimum value if OpenAI’s lease obligations are not met. NVIDIA’s announcement describes residual-value guaranties related to about 4.25 gigawatts of IT load, with the ability to provide further credit support for additional capacity [S1].

Why that matters in practice: an equity check aligns corporate incentives and gives SB Energy runway. The guarantees, meanwhile, act like a lender’s backstop for lease and asset values. They are not the same as handing SB Energy cash to build everything today. They are promises that change the credit profile of the project, making it easier to raise other capital and to sign a large, long-term tenant like OpenAI [S1][S2].

How the lease and tenant commitments work

SB Energy will build, own, and operate the PORTS-Pike data center under a 20-year lease to OpenAI, meaning OpenAI is a long-term tenant responsible for lease payments while SB Energy retains ownership and operational responsibility for the shells and power infrastructure [S2].

That 20-year term is a practical detail that shapes financing. Lenders and equity investors underwrite expected cash flows over long terms. A multi-decade lease to a creditworthy tenant makes the asset financeable, which is where residual-value guarantees enter - they reduce uncertainty about what the owner could recover if a tenant defaulted. In this arrangement OpenAI pays the rent; NVIDIA’s guarantees cover specified residual gaps rather than essentially buying the property outright [S2][S1].

Capacity and timeline: 4.25 GW now, scale to 8+ GW, first capacity in 2028

the project secures roughly 4.25 gigawatts of IT-load capacity under the initial agreements, with documented scope to expand toward 8 IT-GW as additional phases come online, and planned capacity placement beginning in 2028 for the early phases [S1][S2].

Concrete timelines matter because they stagger financial exposure. NVIDIA’s residual-value guarantees become effective as leases commence in each phase. That phasing reduces upfront capital at risk and ties credit support to visible, in-service capacity - lenders can see the asset producing cash before more guarantee tranches kick in [S1].

Operationally, an 8 IT-GW target implies a major build: generation, substations, on-site redundancy, and a sustained hiring pipeline across construction, facilities, and grid-integration roles. The public filings and release language also note a regional power and generation commitment in partnership with local utilities to support the campus scale [S2].

Credit: Photo by Elchinator on Pixabay

How this links back to the earlier Stargate / SB Energy commitments

SB Energy and OpenAI previously announced joint investments to kick-start the developer’s capacity, and this NVIDIA step both increases capital and layers new credit support into the broader Stargate-era commitments. That earlier phase included smaller initial investments and an initial lease footprint the parties expanded on in the August announcements [S2].

Context is important. The January 2026 Stargate announcement set the stage: it signaled strategic investor alignment and gave the site initial credibility. NVIDIA’s follow-on investment adds a supplier-aligned backer and underwriting, which changes the negotiation dynamic with utilities, regional planners, and large-cap contractors. For candidates, that means roles shift from speculative planning to executable delivery.

What the guarantees actually are - and what they are not

public documents describe NVIDIA’s commitments as residual-value guaranties and credit support that back lease obligations for defined portions of the project; they are not described as direct cash grants toward construction costs. That legal and accounting distinction matters for lenders and regulators [S1][S3].

In plain language: a guarantee promises that if lease payments stop, a guarantor covers the difference between an agreed minimum asset value and proceeds the owner recovers from re-leasing or selling. It reduces downside for debt holders and the lessor. It does not mean NVIDIA is handing SB Energy tens of billions in cash up front. Multiple press reports note the headline-sized maximum guarantee figure, but filings make the structural nuance clear: equity plus conditional guarantees, phased as capacity is placed into service [S3][S1][S4].

Why recruiters and hiring managers will care

this deal rewires which teams scale first. Project finance, power engineering, grid-integration, site delivery, and lease operations open first; pure compute deployment teams grow later when shells are ready and racks are sited.

If you are targeting roles in AI infrastructure, know which calendar you want. Construction and power roles start years before racks arrive. Procurement and GPU deployment teams scale as phases hit their bring‑online dates. That means an operations role advertised in 2027 will look different from a compute-deployment role advertised in 2029. (Yes, the hiring timeline can feel like a slow-motion audition - and yes, planning for the right stage matters more than applying to every open role on day one.)

What candidates usually miss

candidates often treat the announcement as a single “job event” and apply broadly, instead of matching their skills to the deal phase and employer need. The 400-applications pattern still happens: high volume, low fit. The smart approach is role-phase matching - pick whether you want to work on power, site delivery, or compute deployment, then prepare accordingly.

Practical specifics: if you want project-finance roles, learn lease structures, residual-value mechanics, and credit support language. If you want grid-integration, focus on interconnection studies, substation design, and utility negotiation. If you want compute deployment, show experience staging at-rack power, fiber handoffs, and large-scale GPU logistics. Those are the exact skills hiring managers will look for when the relevant phase opens.

Frequently Asked Questions

How much is Nvidia investing in SB Energy?

NVIDIA announced a $1.5 billion investment in SB Energy to support SB Energy’s growth and commitments to the PORTS-Pike project [S2]. That is an equity-style backing designed to align NVIDIA with the developer, not a purchase of the entire project.

What is NVIDIA guaranteeing for OpenAI's Ohio data center?

NVIDIA’s public filings and releases describe residual-value guaranties and credit support that back lease obligations for specified capacity - initially related to roughly 4.25 GW of IT load, with the ability to provide additional credit support as phases expand [S1][S3]. The guarantees are structured to become effective as leases commence in each phase rather than as unconditional cash payments to SB Energy.

How large will OpenAI's Ohio data center be?

The initial agreements cover about 4.25 gigawatts of IT-load capacity with scope to scale toward approximately 8 IT-GW as the PORTS-Pike campus adds phases and generation capacity [S1][S2].

When will the OpenAI Ohio data center come online?

Public disclosures point to the first capacity being placed into service beginning in 2028, with additional phases following as the project scales and as power and shell capacity are completed [S1].

How is SoftBank involved in OpenAI's Ohio data center?

SB Energy is a SoftBank affiliate and is the project developer and lessor building and operating the PORTS-Pike campus. SoftBank and OpenAI were earlier investors in SB Energy, and NVIDIA’s investment joins that investor group while adding supplier-aligned credit support to the financing structure [S2][S3].

Final Thoughts

Most candidates treat big announcements as one event and miss that the work behind them is a series of different problems requiring different skills. Pick which problem you want to solve - finance, power, site delivery, or compute deployment - and map your experience to that phase. Start by learning one concrete thing relevant to the role you want (for example, how residual-value guarantees change lender covenants, or how utility interconnection windows are scheduled). It is not glamorous work, but it is the work that actually makes these headlines real. If you can do one of those things well, recruiters will notice; if you try to be all of them at once, you will look like everyone else (and no one hires that person). And yes, reading a securities filing to prepare for an infrastructure interview is mildly nerdy, but it also beats explaining to an interviewer why you thought 400 applications was a strategy.

Explore Nvidia and OpenAI roles and where you fit