MARA Stock Jumps as Miner Locks Down 1,200-Acre Texas Power Campus for AI and Bitcoin

MARA Holdings will buy a 1,200+ acre powered site in Matagorda County, TX, unlocking up to 2 GW by Apr 2028 for AI/HPC and Bitcoin mining. Shares jump as power-first strategy advances.

Bitcoin
Cryptocurrency
Regulations
Economy
Because Bitcoin
Because Bitcoin

Because Bitcoin

July 9, 2026

Bitcoin miners are steadily morphing into power developers, and MARA Holdings just put a bold stamp on that pivot. The company agreed to acquire a large, already-powered tract in Matagorda County, Texas—about 90 miles southwest of Houston—positioning itself to control the scarce input both AI and crypto depend on: dependable megawatts.

The deal secures a 1,200+ acre site from HIF USA with staged grid access: up to 1 GW by October 2027 and as much as 2 GW by April 2028. MARA intends to build a computing campus with Starwood Digital Ventures that can host AI/high‑performance compute alongside Bitcoin mining. HIF will retain a minority interest once a computing tenant signs a lease and will continue pursuing its synthetic-fuels initiatives elsewhere in Texas and overseas.

Investors leaned in quickly. Shares gained more than 15% Thursday to $13.87, per Yahoo Finance, lifting the monthly advance above 4%. Year to date, the stock is up over 54% as AI compute demand—and the power needed to fuel it—becomes the dominant narrative for miners.

Here’s the significance: owning power optionality is evolving into the core business. With this transaction, MARA says its total power portfolio, including a previously announced Ohio power plant acquisition, approaches 4.8 GW—an order of magnitude that starts to resemble regional utilities. That scale can unlock two revenue paths: long-dated AI/HPC leases that prize uptime and proximity to capacity, and flexible Bitcoin mining that can monetize off-peak or volatile pricing, potentially participating in demand-response programs when the grid tightens. The ability to dynamically allocate electrons between deterministic AI workloads and opportunistic hash rate can enhance returns if executed thoughtfully.

Execution risk still matters. Interconnection timelines in Texas can be lumpy, procurement markets for transformers and switchgear remain tight, and moving from 1 GW by late 2027 to 2 GW by spring 2028 is ambitious. Capital intensity will test balance-sheet discipline, and the mix between self-mining and third-party leases will influence cash-flow volatility. Credible milestones—interconnect agreements, tenant signatures, and staged energization—will likely matter more than headlines as the market separates builders from storytellers.

Management’s framing aligns with this thesis. MARA’s leadership emphasized that strategically located, scalable, and reliable power will gain value as digital infrastructure demand expands, and that locking in such sites broadens the development pipeline while increasing the ability to support high-performance compute and to capture the value of that power over time. That’s essentially a shift from being a pure price-taker on electricity to an allocator of compute-ready capacity.

Local impact is tangible. Company officials point to thousands of construction and permanent jobs for Matagorda County as development ramps, adding to more than $1.2 billion MARA says it has already invested in Texas. Community and grid stewardship will remain under scrutiny in ERCOT; miners and AI tenants that curtail during peak stress and communicate transparently tend to earn better operating latitude.

What to watch next: lease signings with AI/HPC tenants, clarity on the phasing to 1 GW and then 2 GW, and how MARA balances self-mining versus third-party compute to smooth earnings without giving up upside. If the capacity timelines hold, MARA increasingly looks like a power-and-compute platform that mines Bitcoin—rather than a miner dabbling in power.