Meta is advancing its AI computing infrastructure more aggressively. According to Michael Thomas, founder of Cleanview, a firm that tracks data center deployments, the company has erected several tent-like structures outside its New Albany campus in Ohio to support its data center project.
This approach is not a temporary stopgap. In a media interview last year, Meta CEO Zuckerberg mentioned plans to use weather-resistant temporary structures to house the company’s multi-gigawatt data center facilities. The latest satellite imagery and local permitting documents have now provided the public with a clearer view of the project’s pace and actual scale.
Five large structures have been built over the past several months.
Thomas’s review of the city permitting documents shows that Meta plans to begin construction of five structures, each approximately 125,000 square feet, between April and June 2026. Satellite images shared by Meta on X indicate that all these facilities have now been fully constructed.
Meta refers to these facilities as "rapid deployment structures." According to publicly available information, the core objective of this approach is to compress the construction timeline. Reports indicate that Meta aims to reduce the completion time by approximately half.
The construction approach is similar to that of Tesla and xAI.
This approach evokes comparisons to Tesla’s early efforts in Fremont, California, where temporary production structures were erected in the parking lot to accelerate Model 3 production ramp-up. Meanwhile, approximately 200 megawatts of modular gas turbines have been deployed near the Meta project to supply power to the campus, mirroring xAI’s current common approach to rapid power provisioning.
Amid extended construction timelines for AI data centers and power connectivity becoming a bottleneck, temporary structures combined with on-site power supply are emerging as a practical solution for some tech companies seeking to accelerate the deployment of computing power.
Under high investment pressure, Meta is placing greater emphasis on speed and cost.
The report states that AI chips will be deployed inside these tent-like structures, with the associated equipment potentially valued at billions of dollars. For Meta, faster delivery of data centers means computing power can be put to use in training and inference tasks sooner.
Previously, Meta stated that it plans to increase its data center and other capital expenditures to as much as $145 billion. This significant investment has raised concerns among capital markets. The report noted that Meta’s stock has declined approximately 5% year-to-date. In this context, adopting faster and potentially lower-cost construction methods is seen as a way to alleviate spending pressure.
As of the time of this report, Meta has not responded to the matter.