The Zhitong Finance App learned that as more and more cryptocurrency mining companies migrate power access, computer room construction, and high-density computing power operation capabilities to AI data center businesses, BTC Digital (BTCT.US) is becoming the most typical example of transformation from a simple cryptocurrency mining company to a “cryptocurrency computing power resource+AI computing power infrastructure resource provider”.
The core logic that its stock price surged close to 60% before the market on Monday is undoubtedly that the 10-megawatt project in Georgia in the US moves from the construction phase to the actual deployment and potential AI inference computing power revenue realization stage. The maximum theoretical computing power limit is about 900,000 TH/s; however, whether the valuation can continue to expand still depends on the actual deployment scale, electricity progress, mining machine efficiency, and digital asset prices, rather than the theoretical computing power itself.
Digital computing power infrastructure company BTC Digital said in a pre-market announcement on Monday that its cryptocurrency computing power infrastructure project in Georgia has been completed and is currently close to being “ready for deployment”, driving the stock to skyrocket nearly 60% during Monday's pre-market trading.
Based on the current progress of electricity access and pre-deployment preparations, the cryptocurrency mining company is expected to deploy a high-performance cryptocurrency miner and launch a digital asset mining business within about two months.
The 10-megawatt facility is designed to have a maximum theoretical computing power of about 900,000 TH/s, provided that all the energy-efficient mining rigs that BTC Digital plans to configure are deployed. The actual number of cryptocurrency mining machines deployed, the final level of computing power, and the pace of operation will be adjusted according to factors such as power-up progress, mining machine model and energy efficiency, network difficulty, digital asset prices, and overall market conditions.
At the same time, the company will continue to develop its comprehensive AI computing power operation service capabilities at the AI computing power infrastructure level. BTC Digital's management said in a statement: “We will continue to maintain the development process of the cryptocurrency computing power business to give full play to the company's existing advantages in power resources, computing power infrastructure leasing and comprehensive computing power operations around extremely scarce AI GPU computing power devices, and build a more diversified and resilient business foundation for long-term growth.”
Bitcoin (BTC-USD) traded up about 1.0% to around $78,500 during pre-market trading in the US stock market on Monday.
From Bitcoin mining machines to AI data centers, the wave of transformation of mining companies to AI data centers is about to add more samples
Currently, BTC Digital is still a digital computing power infrastructure company with Bitcoin mining as the main cash flow. Its main business includes cryptocurrency mining, mine computing power construction and operation, and related GPU computing power infrastructure services; the maximum design theoretical computing power of the 10 MW project in Georgia, USA is about 900,000 TH/s. The company also plans to use land, electricity and construction resources in the same park to build the first 8 megawatt AI computing center. It plans to use a wholesale hosting model - BTCT to provide electricity, computer rooms, networks and operation and maintenance. Customers bring their own GPUs and bear hardware depreciation.
However, the company clearly disclosed to the US SEC that its AI business is still in its early stages and has yet to obtain binding customer contracts or generate AI revenue, so the 54% pre-market surge mainly reflects the option value of “power assets converted into AI revenue” rather than profits that have already been realized.
Bitcoin miners are moving to the underlying logic of AI data centers, where two types of businesses share the most scarce production factors: grid-connected power (Grid-Connected Power), land, substations, computer room housings, cooling infrastructure, and 24/7 computing power operation and maintenance capabilities. Compared to mining revenue, which is dominated by currency prices, full network difficulty, and halving cycles, AI hosting or GPU cloud services can generate more stable recurring revenue through multi-year contracts; especially in environments where grid access takes several years, the “time-to-power advantage” (time-to-power) that miners already have is a high-value asset.
However, the transformation is not simply replacing ASIC miners with GPUs: AI data centers must also complement highly reliable power supply, optical fiber networks, direct touch chip liquid cooling, network operation activity orchestration, security systems, and service level agreements (SLAs). The reason why cryptocurrency mining company IREN is a successful sample is because it has vertical integration capabilities from site selection and engineering to GPU cloud operations; its first 50 MW Horizon facility has been tested by Microsoft and certified by Nvidia Exemplar Cloud. The company aims to develop 480 megawatts of AI cloud capacity in 2026 and 1.2 gigawatts in 2027.
These transformation to data centers or mining companies that are transforming their valuation anchors are shifting from currency prices and unit hash rates to megawatts of electricity, contract backlog, GPU utilization, annualized recurring revenue (ARR), and return on capital. IREN expects AI Cloud ARR to exceed US$4 billion by the end of 2026, and approximately 85% is supported by contracts, indicating that the market is willing to give cloud infrastructure premiums for “long-term locked token production capacity.”
However, BTCT is currently closer to miners with AI transformation options rather than the proven IREN-style AI cloud platform. Whether it can be continuously revalued depends on customer anchoring, closed loop financing, liquid cooling transformation, on-time delivery, and real AI revenue generation; if these steps are not realized, the current premium may still degenerate into an event-driven market for small-cap stocks.