Galaxy Digital and Mara Holdings have each secured new sites in the U.S. state of Texas as they move to expand infrastructure for artificial intelligence (AI), high-performance computing (HPC) and bitcoin mining.
Cointelegraph, a blockchain media outlet, reported on July 28 local time that the announcements came as demand continues to grow for power-hungry AI and digital infrastructure.
Galaxy Digital said it bought a 500-acre site in McGregor, Texas. It plans to develop its second AI and HPC data center campus in the state there. The project will start with 74 megawatts of power capacity and then expand in phases.
Mara Holdings announced it signed a contract to acquire a 1,200-acre, power-connected site in Matagorda County. The site can access up to 2 gigawatts of power capacity. The company said it plans to develop the site for AI and HPC work as well as bitcoin mining.
It is also notable that the two companies unveiled plans for large infrastructure investments in Texas on the same day. Meta Platforms also unveiled plans that day to build a $14 billion AI data center campus in El Paso. Competition for large-scale digital infrastructure investment in Texas appears to be accelerating further.
The reasons Texas has emerged as a hub for such investment are also clear. Texas is viewed as having a large power market, relatively low electricity rates and land suitable for large developments. The state’s independent power grid, ERCOT, is also cited as a factor attracting developers. It is a key condition in site selection for operators seeking access to large power capacity.
Against this backdrop, the expansion of crypto companies in Texas is difficult to view as merely adding mining facilities. As power-securement-centric AI data centers and bitcoin mining infrastructure intersect in the same region, access to land and the power grid is emerging as a key variable in business competitiveness. Attention is focusing on how much more investment Texas will attract as a base where AI and digital asset infrastructure converge.