오프그리드 AI 전력 회사 TAR이 Spark Capital 주도로 시리즈 A 펀딩에서 $120 million을 조달하며 $1 billion의 포스트머니 밸류에이션에 도달했다.

Power availability is rapidly becoming a primary constraint on AI data center expansion. Grid interconnection timelines and transmission limitations are slowing projects even when computing hardware is available. TAR is addressing this constraint through an off-grid approach, having raised $120 million in Series A funding to scale dedicated power systems for AI infrastructure.

Spark Capital led the financing, valuing the Austin-based company at $1 billion post-money. The capital will fund expansion of TAR's Austin headquarters and San Francisco engineering office, increased manufacturing and logistics capacity in West Texas, and acceleration of power projects currently under development.

Founded in 2026 by Pat Becker and Leonhard "Lenny" Soenke, TAR develops self-contained power systems combining renewable energy generation with battery storage, designed to supply AI data centers without relying on conventional grid interconnections. The company targets a critical infrastructure gap: large computing campuses require substantial electricity, yet connecting new projects to the grid often involves multiyear queues and capacity constraints on generation and transmission.

TAR is shortening deployment timelines by controlling the entire energy development process internally—site selection, engineering, design, procurement, logistics, civil construction, commissioning, and ongoing operation—rather than coordinating across multiple outside providers. "To lead the frontier of AI after a lack of investment in the electrical grid for decades requires a complete rethinking of energy deployments," Becker said. "Gigawatt-scale deployments in tight time windows necessitate owning the full stack end-to-end."

The company applies automation to physical energy deployment through purpose-built technology intended to reduce field labor and accelerate infrastructure installation. Its team combines conventional power industry experience with robotics and advanced manufacturing backgrounds, drawing employees from energy companies including Hut 8, AES, and Vistra, alongside technology and robotics businesses including Zipline, GrayMatter Robotics, and Lucid Motors.

TAR is currently executing a utility-scale deployment with one of the largest neocloud operators while developing a dedicated project campus and completing TAR Terminal One, its West Texas manufacturing and logistics facility. Leadership includes Jeff Silvan (head of projects, formerly Hut 8 and ON Energy), Christian Sanchez (head of power systems, previously Zelestra and AES), and Raphael Levy (head of supply chain, formerly NineDot and Vistra).

"Our focus now is scaling supply to make a real impact on the power shortage stopping the scaling of compute," Soenke said. The Series A funding targets physical deployment capacity rather than technology development alone, reflecting TAR's strategy of building manufacturing, logistics, and construction capability alongside expanding AI compute campuses.

For AI infrastructure developers, off-grid power offers a route around one of the longest lead-time components of data center development, shifting responsibility for generation, storage, and reliability into the data center project itself rather than depending entirely on utility availability. TAR's $1 billion valuation reflects investor recognition that GPU access alone no longer determines how quickly capacity comes online—vertically integrated energy development, automated construction, and dedicated renewable generation have become critical infrastructure enablers.