Datacentrex (NASDAQ:DTCX) is currently running more than 3,100 Scrypt mining rigs across four colocation sites in the United States. The company has disclosed that these facilities are live, powered, and operating today with uptime above 95%. Revenue is being generated under continuous load.
Management has announced a funded expansion to more than 4,100 rigs in the first half of 2026. This growth will occur inside facilities that are already running. Power, cooling, and operating processes are in place. Capital is directed toward machines, not construction.
The company's Scrypt operations have been EBITDA positive from inception, according to company materials. Power costs are covered. Sites fund themselves. Datacentrex has disclosed roughly $45 million on hand, no long-term debt, and full ownership of its operating equipment.
Scrypt mining secures networks such as Litecoin and Dogecoin. It is a different market from Bitcoin mining. The field is smaller. Competition is thinner. Scale is not dominated by a handful of megasites.
Company materials show that Scrypt mining delivered stronger realized profitability than SHA-256 mining for much of 2024. CoinMetrics network data show that total miner revenue on Scrypt networks such as Dogecoin and Litecoin stayed comparatively steady through 2024.
Scrypt mining allows a single machine to earn rewards from two networks at the same time, most commonly Litecoin and Dogecoin. This feature is known as merged mining. Power use does not increase, but output does.
Datacentrex handles mining output by consolidating proceeds into bitcoin rather than managing several smaller assets. The result is a simpler cash flow. Fewer conversions. Less exposure to liquidity bottlenecks. Revenue moves from production to cash without detours.
The company owns its machines and operates directly within third-party colocation facilities. There are no leased rigs and no vendor liens embedded in the fleet. Decisions around uptime, expansion, and capital allocation stay simple.
Microsoft (NASDAQ:MSFT) has concentrated new AI capacity inside existing campuses in places like Iowa and Texas, where power delivery and transmission are already settled. Google (NASDAQ:GOOG) has done the same in long-established regions such as Virginia and Ohio. These are not speculative locations.
J.P. Morgan estimates that U.S. data center capacity could exceed 130 gigawatts by 2030, up from roughly 20–30 gigawatts today. Total infrastructure and hardware investment required to support that buildout could run north of $3 trillion.
J.P. Morgan’s work shows that U.S.-listed bitcoin miners are operating more than 8.5 gigawatts of energized power, with approvals in place for roughly another 8 gigawatts. The sites are connected, permitted, and already drawing load.
J.P. Morgan estimates that converting existing mining power capacity to HPC applications could cost roughly $10-15 million per megawatt before GPUs. All-in AI data center costs could exceed $45 million per megawatt once hardware is included.
Datacentrex is not trying to solve the power problem. It is operating after it. Its business decisions concern how to use existing capacity, not how to secure it.