Google Acquires Crusoe’s Wyoming Data Center Project

Table of Contents

Google invests in troubled Wyoming data center project

  • Crusoe confirmed the sale, according to TechCrunch.
  • Google said it plans to invest $17 billion to develop the campus, according to TechCrunch.
  • Bloomberg reported Google raised concerns about cost and construction timelines under Crusoe’s management.
  • Crusoe is still building another data center for Google near Amarillo, Texas, according to TechCrunch.

Methodology

This article is based on reporting and disclosures from TechCrunch, Bloomberg, and Google, plus project-level details referenced by ServerCountry and local reporting cited in the broader coverage. The core transaction was confirmed by Crusoe, according to TechCrunch.

To avoid overstating what is known, the piece distinguishes between (1) confirmed statements from the companies involved, (2) details attributed to specific reporting outlets, and (3) project specifications attributed to project trackers and public filings as described by those outlets. For example, TechCrunch reported Google did not disclose the purchase price, while Google publicly announced an investment plan for the campus, as reported by TechCrunch and described by Google’s own blog.

Where motivations are discussed—such as Google’s concerns about schedule and cost—those points are attributed to Bloomberg’s reporting. Where the broader market context is addressed—such as the difficulty independent developers face in delivering massive projects on time and within budget—the framing follows TechCrunch’s analysis of the deal dynamics, including equipment, power, and community-opposition constraints.

Overview of Crusoe’s Wyoming Data Center Project

Crusoe, an eight-year-old data center developer based in Denver, set out to build a major data center campus in southeastern Wyoming in partnership with Tallgrass, a Blackstone-backed infrastructure company, according to TechCrunch. The ambition was enormous: the project was planned to eventually scale to 10 gigawatts of power capacity, TechCrunch reported.

In later project descriptions compiled by ServerCountry, the site is described as a 716-acre parcel in the Switchgrass Industrial Park south of Cheyenne in Laramie County, with a multi-building campus plan and phased power delivery. ServerCountry also described revisions that reduced the number of buildings and lowered peak workforce estimates compared with earlier plans.

The Wyoming effort mattered because it sat at the intersection of two constraints that increasingly define AI-era infrastructure: access to power and the ability to build quickly. TechCrunch noted that independent developers face major challenges securing equipment and power, and that growing opposition from local communities can complicate development. Those pressures are amplified at gigawatt scale, where delays in grid interconnection, generation buildout, or supply chains can cascade into missed delivery windows for customers that are racing to deploy AI compute.

Crusoe’s corporate arc also helps explain why it pursued projects of this magnitude. TechCrunch reported the company was founded by Chase Lochmiller (now CEO) and Cully Cavness, and that it began as a cryptocurrency-mining business powered by natural gas that would otherwise have been burned off.

Google’s Acquisition of the Wyoming Project

Crusoe confirmed it sold the unfinished Wyoming data center project to Google, according to TechCrunch. The sale is notable not only because it transfers a large, partially developed campus to a hyperscaler, but because Google had been a target customer for the project and Crusoe failed to land it in that role, TechCrunch reported.

Bloomberg first reported the deal in June, according to TechCrunch, and Bloomberg later described Google’s takeover as being driven by concerns about cost and construction timelines. In other words, Google moved from prospective customer to owner-operator—an escalation that changes who bears execution risk and who controls delivery.

Google did not disclose how much it paid Crusoe for the unfinished project, TechCrunch reported. What Google did disclose was its intent to spend heavily going forward: Google announced plans to invest $17 billion to develop the data center campus, according to TechCrunch. Google also described the Wyoming investment on its corporate blog, which TechCrunch characterized as a major commitment.

The transaction underscores a structural shift in the market: when a project becomes strategically important to a hyperscaler’s AI roadmap, ownership and control can matter as much as the real estate itself. Bloomberg’s reporting framed Google’s move as a response to execution concerns, while TechCrunch emphasized the broader difficulty for independent developers to deliver massive builds on time and within budget.

Concerns Raised by Google Regarding Project Management

Google’s concerns centered on two classic failure modes in megaproject delivery: cost and time. Bloomberg reported that Google raised concerns about the project’s cost and construction timeline under Crusoe’s management, and that Google ultimately took over development.

That matters because hyperscale data centers are no longer “just” real estate and servers; they are power-hungry industrial systems with long-lead equipment, complex permitting, and tight coupling to generation and transmission plans. TechCrunch highlighted the practical obstacles: securing equipment and power can be difficult, and local community opposition can further complicate development. Each of those variables can push schedules out and budgets up—especially when the target is gigawatt-scale capacity.

Bloomberg’s reporting also adds an important nuance: Crusoe had tried and failed to land Google as a customer for the Wyoming project, TechCrunch reported, and Google’s later takeover suggests the company preferred direct control rather than relying on Crusoe to deliver a facility that met Google’s requirements on price and timing.

In the AI infrastructure race, timing is not a soft metric. If a campus is late, the downstream impact can include delayed model training capacity, slower product rollouts, and higher costs for interim compute. Bloomberg’s account—concerns about cost and timeline—fits that reality without requiring any additional speculation about internal decision-making.

Investment Plans for the Data Center Campus

Google announced plans to invest $17 billion in the Wyoming data center campus, according to TechCrunch. The company did not disclose what it paid Crusoe for the unfinished project, TechCrunch reported, but the forward-looking investment figure signals that Google intends to push the campus toward full hyperscale operation rather than treating it as a minor regional site.

Project trackers and reporting cited in the broader coverage provide a sense of the campus’s intended scale. ServerCountry described the project as having been planned with multiple buildings and phased power delivery, and noted revisions that reduced the number of buildings and peak workforce estimates. While those specifics come from a project compilation rather than a single corporate press release, they align with the broader point TechCrunch made: delivering massive data center projects is hard, and plans often change as developers confront real-world constraints.

The $17 billion figure also illustrates a key asymmetry in the market. Independent developers can assemble land, entitlements, and early infrastructure plans, but hyperscalers can often deploy capital at a level that allows them to absorb overruns, lock in supply chains, and standardize construction and operations across a global fleet. TechCrunch framed the deal as a case study in how difficult it can be for independent developers to deliver massive projects on time and within budget.

At the same time, Google’s investment plan does not mean the project becomes simple. The same constraints TechCrunch cited—equipment, power, and community dynamics—still apply. The difference is that Google is now the party directly managing those risks.

Crusoe’s Financial Resilience Despite Setbacks

The Wyoming setback did not stop Crusoe from raising significant capital. TechCrunch reported that Crusoe raised a $3.9 billion round in September, after Bloomberg first reported the Wyoming deal in June. TechCrunch also reported that Crusoe was recently valued at $30.9 billion.

That ability to raise money despite losing control of a flagship project suggests investors still see Crusoe as a meaningful AI infrastructure player—likely because the company is not dependent on a single campus. TechCrunch described Crusoe’s business model as multifaceted: it leases data center space to customers that bring their own GPUs, rents out its own GPUs, and sells compute power used to run AI models (inference).

Crusoe’s customer roster and pipeline of large builds also help explain the resilience. TechCrunch reported that Crusoe’s campus in Abilene, Texas, is being built in phases for Oracle and OpenAI. TechCrunch also reported Crusoe is constructing a separate 900-megawatt campus next door for Microsoft, with its first building expected to come online in mid-2027. Meta is also a customer, TechCrunch reported.

Bloomberg also reported that trading firm Jane Street signed a $13 billion, five-year cloud deal with Crusoe, according to TechCrunch. Taken together, those projects and customers indicate that Wyoming was a setback—but not necessarily an existential one.

Future Developments: New Data Center Near Amarillo, Texas

Even after taking over the Wyoming facility, Google has not walked away from working with Crusoe. TechCrunch reported Crusoe is developing another data center for Google in Texas.

That detail is important because it complicates any simplistic narrative of “Google lost faith in Crusoe.” Instead, the reporting suggests a more specific conclusion: Google was not satisfied with Crusoe’s management of the Wyoming project’s cost and timeline, as Bloomberg reported, but still sees value in Crusoe as a developer and compute partner in other contexts.

From an execution standpoint, it is also plausible that different sites carry different risk profiles—power availability, permitting complexity, supply chain logistics, and local stakeholder dynamics can vary widely. TechCrunch emphasized that securing equipment and power is challenging and that local opposition can complicate development; those factors can be more acute in some jurisdictions than others.

The Amarillo project also reinforces the idea that AI infrastructure is being built through a mix of ownership models. In one case, Google is taking direct control of a campus. In another, it is still relying on a specialist developer. The common thread is urgency: the demand for AI compute is pushing companies to pursue multiple parallel paths to capacity.

Crusoe’s Role in the AI Infrastructure Market

Crusoe sits in a fast-evolving segment where the lines between developer, operator, and compute provider are blurring. TechCrunch reported that Crusoe makes money by leasing data center space to customers that bring their own GPUs, renting out its own GPUs, and selling inference compute. That combination positions Crusoe not only as a landlord, but as a compute supplier—an important distinction in an era when access to GPUs and power can be as strategic as access to real estate.

TechCrunch also reported that Crusoe’s largest projects serve some of the biggest names in AI: Oracle and OpenAI in Abilene, Microsoft in a neighboring 900-megawatt campus, and Meta as a customer. Those relationships place Crusoe in the middle of the AI buildout, even as hyperscalers increasingly prefer direct control over certain assets.

Corporate governance and strategic partnerships also feature in Crusoe’s story. TechCrunch reported that alongside the $3.9 billion funding round in September, Crusoe added three board members, including JB Straubel, the Redwood Materials founder and Tesla board member who invested in Crusoe in 2021. TechCrunch also reported Crusoe later became the first customer of Straubel’s energy-storage business.

That detail matters because power and energy storage are increasingly central to data center economics and reliability. While TechCrunch did not quantify how storage is deployed in Crusoe’s projects, the connection underscores the broader theme: AI infrastructure is as much an energy and industrial-systems challenge as it is a software or cloud challenge.

Conclusion: The Future of Data Centers and AI Infrastructure

Author context

Weidemann.tech is written by Martin Weidemann, a serial entrepreneur and digital transformation strategist with hands-on experience building technology-driven businesses in regulated environments, including payments and infrastructure-adjacent systems. This background informs how the article frames execution risk, timelines, and operational constraints in large-scale technology projects, while all deal-specific facts and figures above are attributed to the cited reporting.

The Impact of Google’s Acquisition on the Market

Google’s takeover of the Wyoming project highlights a market reality TechCrunch put plainly: it can be extremely difficult for independent data center developers to deliver massive projects on time and within budget, especially amid challenges securing equipment and power and navigating local opposition. Bloomberg’s reporting adds a sharper edge—when a hyperscaler is unhappy with cost and timeline risk, it may choose to take control rather than remain a customer.

Google’s announced $17 billion investment plan, reported by TechCrunch, signals that hyperscalers are willing to spend at extraordinary levels when they believe a site can become strategically important capacity. The deal also suggests that “development rights and entitlements” can be valuable even when a project is unfinished—because they compress the time needed to reach build-ready status.

For the broader AI infrastructure market, the implication is not that independent developers disappear. Crusoe’s continued work for Google near Amarillo, reported by TechCrunch, and its major builds for Oracle, OpenAI, Microsoft, and Meta, also reported by TechCrunch, show that specialist developers can still play a central role. But the balance of power is clear: hyperscalers can—and will—step in when execution risk threatens their timelines.

Lessons Learned for Future Data Center Developments

The Wyoming episode offers a few grounded lessons drawn from the reporting:

First, anchor customers are everything. TechCrunch reported Crusoe tried and failed to land Google as a customer for the Wyoming project; Bloomberg reported Google had concerns about cost and timeline. Without a committed hyperscale tenant, financing and execution become harder, and the developer’s leverage shrinks.

Second, power and equipment constraints are not background noise—they are core project risks. TechCrunch emphasized the challenges of securing equipment and power, and noted that local community opposition can complicate development. At gigawatt scale, those constraints can force redesigns, delays, and ownership changes.

Third, setbacks do not necessarily end a company’s story in this market. TechCrunch reported Crusoe raised $3.9 billion at a $30.9 billion valuation even after the Wyoming deal was first reported by Bloomberg. That resilience reflects both investor appetite for AI infrastructure exposure and Crusoe’s broader portfolio of large customers and campuses.

Finally, the industry is moving toward a hybrid model: hyperscalers will own and operate some campuses directly, while still partnering with developers like Crusoe on others. The common denominator—visible in TechCrunch’s reporting across customers, campuses, and capital raises—is that AI compute demand is pulling the entire supply chain forward, even when individual projects stumble.

Related Posts

Scroll to Top