Bitcoin Miner MARA Buys Ohio Power Plant for $1.5 Billion, Pivots to AI Infrastructure
MARA Holdings signed a definitive agreement to acquire Long Ridge Energy & Power from FTAI Infrastructure for approximately $1.5 billion. The deal includes a 505 MW combined-cycle gas power plant in Hannibal, Ohio, plus 1,600 contiguous acres already equipped for large-scale computing facilities.
The acquisition boosts MARA's owned-and-operated capacity by 65%, bringing its total to roughly 2.2 GW. The company projects the Long Ridge asset will contribute approximately $144 million in annualized adjusted EBITDA, with operational costs below $15 per megawatt-hour.
Why a Bitcoin Miner Needs a Power Plant
Bitcoin mining margins compressed sharply after the April 2024 halving cut block rewards in half. MARA's move reflects an industry-wide shift: competitors including Core Scientific, Iris Energy, and Hut 8 are all pivoting toward high-performance computing and AI infrastructure contracts.
The difference in MARA's strategy is ownership. Rather than buying electricity on the open market, MARA now owns the power source itself. That sub-$15/MWh operating cost becomes a competitive advantage when bidding for AI hosting contracts.
MARA described the $1.5 billion price as "below-replacement cost," suggesting equivalent new construction would cost significantly more.
Timeline and Investor Considerations
Construction for AI and critical IT infrastructure at Long Ridge begins in the first half of 2027, with initial capacity targeted for mid-2028. That 18-month wait means near-term value depends on the power plant's existing cash flows and EBITDA contribution.
The sub-$15/MWh operating cost is the metric to track. It determines whether MARA can undercut competitors on infrastructure hosting rates and win contracts at scale.
For IT and development professionals evaluating infrastructure providers, MARA's vertical integration of power generation and computing capacity represents a structural shift in how AI compute pricing may be negotiated. Understanding AI for IT & Development requires knowing how infrastructure economics shape deployment options and costs.
Your membership also unlocks: