Meta steals a tactic from Tesla and builds data centers in tents
Original reporting by TechCrunch
The relentless expansion of AI infrastructure has taken a dramatic, even unconventional, turn. Just when the scale of the AI data center boom seemed to defy further escalation, Meta has begun deploying massive artificial intelligence computing power within "rapid deployment structures"—effectively, tents. This audacious strategy, identified through satellite imagery and local permits by data center tracker Michael Thomas, reveals six immense structures outside New Albany, Ohio, with permits indicating five 125,000-square-foot units rapidly erected between April and June 2026. This initiative is designed to slash construction times by half, a critical move in the fiercely competitive AI race.
Strategy and urgency
This unprecedented approach echoes Tesla's use of temporary facilities for Model 3 production, while the 200 megawatts of modular gas turbines powering the site recall competitor xAI's innovative off-grid tactics. Meta’s urgency is palpable: The company has faced internal delays in releasing its latest AI models to developers due to API bottlenecks, exacerbating the pressure to quickly expand its computational muscle. Moreover, Wall Street has reacted poorly to Meta’s projected $145 billion capital expenditure on data centers, pushing its stock down. Housing billions of dollars in AI chips within these expedited, unconventional facilities offers a clear path to trim both time and expense, signaling that the AI infrastructure buildout has indeed entered its "Mad Max phase."
Meta's deployment of tent-based data centers in Ohio represents a radical, yet increasingly necessary, evolution in AI infrastructure. Driven by an insatiable demand for compute power and the urgent need to accelerate AI model development, this strategy underscores the intense pressures facing tech giants in the current AI arms race. By embracing modular, rapidly deployable structures and unconventional power sources, Meta is not just building data centers faster; it is rewriting the playbook for scaling foundational AI capabilities under extreme conditions. This bold move, reminiscent of past industrial surges, highlights a willingness to innovate aggressively to overcome traditional bottlenecks in construction, supply chains, and energy provision.
A new paradigm
The broader implications of Meta's "rapid deployment structures" are profound, signaling a potential shift towards more agile, less conventional infrastructure across the technology sector. This approach could normalize the use of temporary or semi-permanent facilities, challenging traditional notions of data center permanence and design. For the AI industry, it means the pace of innovation could further accelerate, as the time from conception to operational capacity dramatically shrinks. However, it also raises questions about long-term sustainability, security, and the integration of such facilities into existing energy grids. Future impact could see competitors adopting similar strategies, further intensifying the demand for modular components and off-grid power solutions, ultimately reshaping the landscape of global AI infrastructure into one defined by speed, adaptability, and an unprecedented focus on raw compute output at any cost.