Bitdeer Masters Megawatt Infrastructure for AI Transformation Game

On February 20, 2026, Bitdeer released a weekly production update that caught the market’s attention. The company successfully mined 189.8 BTC that week but sold it immediately. The remaining Bitcoin stock of 943.1 BTC was also sold off at once, leaving the company’s Bitcoin balance at zero. This drastic decision was purposeful—the proceeds from the sales were used to secure global infrastructure assets, especially massive-capacity power lines. This move marked a turning point in Bitdeer’s business strategy: shifting from a coin-mining company to a large-scale energy infrastructure player for the AI era.

Megawatt Units: The True Measure of Bitdeer’s Infrastructure Ambition

To understand the scale of Bitdeer’s ambitions, we need to grasp what a megawatt (MW) is. MW is a unit of electrical power equal to one million watts—how the energy industry measures generation and distribution capacity. When talking about data centers and energy-intensive facilities, megawatts are a key metric. Each MW of operational capacity represents a continuous energy flow capable of supporting thousands of computing devices.

By early 2026, Bitdeer had accumulated a global power capacity of 3,002 MW, with 1,658 MW operational and 1,344 MW still under construction or awaiting activation. These numbers may sound abstract until we compare them to industry standards. Major data centers owned by Microsoft and Google—two leading tech giants—typically have capacities of 100 to 300 MW per location. This means Bitdeer’s total capacity is equivalent to consolidating the energy needs of 10 to 30 massive data centers within a single corporate entity.

The question is: how did Bitdeer achieve this scale? The answer lies in a decade of experience running Bitcoin mining operations. This decade not only generated profits but also granted access to premium electrical infrastructure worldwide—assets now foundational to the transformation into AI data centers.

From Bitcoin Mining to AI Data Centers: The Same Arbitrage Logic Over Time

In the mining world, there’s a fundamental logic that has persisted for the past twelve years: using electricity and machines today to earn Bitcoin tomorrow. This business model is simple yet powerful—no factories needed, no traditional customers, no brand required. The investment is in current energy costs; the wager is on the future value of the asset.

Now, under Wu Jihan’s leadership, Bitdeer is revolutionizing this logic without changing its core foundation. The focus shifts from Bitcoin’s price to the long-term value of computational power in the AI era. The method evolves from consuming electricity to mine coins into borrowing capital to secure land and energy assets. Yet, the essence of arbitrage over time remains the same: buy a position today at a certain cost, hoping its value will increase dramatically in the future.

This transformation requires substantial capital. In May 2024, Tether invested $100 million, becoming the second-largest shareholder. Three months later, the first round of $150 million convertible bonds was completed at an 8.5% annual interest rate. By November 2024, a second convertible bond of $360 million was issued at a lower rate of 5.25%. The momentum continued: in November 2025, a combined offering of $400 million in convertible bonds and an additional $148.4 million in equity was announced. By February 2026, Bitdeer issued another $325 million in convertible bonds while refinancing $135 million of old debt, extending maturities to 2032.

Total capital raised exceeded $1.4 billion in less than two years. The funds flowed into computing devices, physical data centers, and AI infrastructure across various countries. Each new bond issuance triggered market reactions, with stock prices dropping 10-17% as a sign of investor skepticism. Nonetheless, Bitdeer successfully secured financing, indicating that some investors believe in this long-term transformation narrative.

Megawatt Projects: Execution Pathways and Critical Risk Points

The raised funds are allocated to three main infrastructure projects that form the backbone of the strategy:

Rockdale, Texas – 563 MW (including a 179 MW expansion). Fully operational, this facility focuses on Bitcoin mining with stable cash flow. It supports Bitdeer’s short-term profits, generating predictable revenue while other infrastructure projects are in setup stages.

Clarington, Ohio – 570 MW. This is the core of Bitdeer’s AI transformation. A 30-year power contract has been signed, with completion initially scheduled for Q2 2027. This facility will serve as an HPC (High-Performance Computing) and AI hub, no longer just a mining center. However, this project also carries the greatest current risk—and will be discussed in more detail.

Tydal, Norway – 175 MW. The most advanced and low-risk project, Bitdeer is converting an existing Bitcoin mine into an AI data center, leveraging local hydroelectric resources. The effective IT capacity reaches 164 MW, with completion scheduled for late 2026. The transformation cost is much lower than building new infrastructure from scratch.

These three assets—land, power lines, and server space—are what the AI industry calls “hard-to-imitate assets.” Bitdeer has accumulated them through a decade of mining operations. An important point often overlooked: Bitdeer is not only building facilities but also developing its own mining chips under the SEALMINER brand.

The SEAL series has reached its third generation, with the SEAL03 achieving an energy efficiency of 9.7 joules per terahash—top of its class. The A3 Pro, which began mass production in September 2025, is already among the top global models. SEAL04 aims for an efficiency of 5 joules per terahash; if achieved, it will surpass all mass-market mining machines. The gross margin of self-developed chips exceeds 40%, far higher than traditional mining margins. This echoes Wu Jihan’s previous approach at Bitmain: from buying chips from suppliers to manufacturing their own.

$1.3 Billion Debt and a Narrow Time Window

Bitdeer’s financial burden is very real. As of December 31, 2025, its book debt reached $1 billion; combined with the new $325 million convertible bonds settled in February 2026, total debt exceeds $1.3 billion.

With an assumed average interest rate of 5% on $1.3 billion principal, annual interest expenses are over $65 million. Meanwhile, revenue from the AI/HPC Cloud business in 2025 was less than one-tenth of the interest burden in six months—meaning this new business has yet to generate significant revenue to cover financial costs.

Currently, these interest payments rely entirely on continuous debt issuance to keep operating. The pressure is substantial. However, Bitdeer’s debt structure is carefully designed. The company set maturities for three series of convertible bonds in 2029, 2031, and 2032—creating what can be called an “execution buffer window.”

The logic: when the first batch matures (2029), the Tydal and Clarington facilities should be operational; by the second batch’s maturity (2031), AI revenue should be significantly measurable; by 2032, the market will assess whether this transformation has succeeded. These three maturities are three opportunities for renegotiation or equity conversion.

However, Wall Street is not fully convinced. Keefe Bruyette lowered its price target from $26.50 to $14. The current stock price hovers around $8. Market signals are very realistic: the transformation story must be supported by concrete revenue growth, not just infrastructure ambitions.

Best-Case Scenario and Risk Traps

In the best-case scenario, the timeline might look like this: by late 2026, Tydal’s renovation is complete, operating fully as a hydroelectric AI data center in Norway, with European customer contracts underway. In 2027, legal issues at Clarington are resolved, construction of the 570 MW Ohio project officially begins, followed by major US clients. From 2028 to 2029, both core assets reach full operation, revenue climbs into the billions, and market analysts rebrand Bitdeer from a discounted mining company to a premium AI infrastructure player.

But there’s a serious trap to watch out for: Clarington.

In the same Ohio industrial park, there’s a steel producer called American Heavy Plate Solutions, which signed a 30-year lease for 9.9 acres in 2018. They sued Bitdeer, claiming that building the AI data center would disrupt power flow, access roads, rail lines, and shared communication lines—violating lease restrictions. Their simple demand: a court-issued injunction to permanently block Bitdeer from starting construction.

Clarington accounts for 42% of the ongoing power line pipeline. If this project is delayed, Bitdeer’s entire execution schedule must be rewritten. The biggest single risk currently isn’t debt or stock price but a steel plant in Ohio.

In Bitcoin mining, pressure also continues to mount. In February 2026, Bitcoin network difficulty surged 14.7%—the largest single increase since May 2021. With the same computational capacity, the amount of Bitcoin mined decreased. Q4 2025 gross margins fell from 7.4% a year earlier to 4.7%. Traditional mining operations are gradually becoming thinner.

The worst-case scenario is clear: a two-year delay in Clarington’s lawsuit, halted construction; Tydal experiencing delays, GPU utilization stagnating at 41%; the first bonds maturing in 2029, with insufficient cash on hand, forcing refinancing, stock dilution, and increasingly difficult conversion thresholds.

Infrastructure as Strategy: Controlling Power Lines, Not Guessing Outcomes

In the mining industry, there’s a tradition of hoarding coins—holding Bitcoin as a sign of confidence in its long-term value. Marathon Digital hoards 53,250 BTC, Riot 18,000, and Strategy 710,000. The more coins are hoarded, the more the market believes in the company’s long-term commitment.

Now, Bitdeer’s balance sheet shows zero Bitcoin. The official explanation: selling Bitcoin to provide liquidity for infrastructure asset purchases. This is true. Other companies are moving in the same direction: Riot sold $200 million worth of Bitcoin to expand into AI; Bitfarms is shedding its “pure-play Bitcoin” identity; Marathon is also building HPC capacity.

But there’s a deeper shift at play. Since the beginning, the mining industry has always wagered one thing: that an asset in the future will be worth more than the current cost. Ten years ago, the industry bet on Bitcoin’s appreciation. Now, buying power infrastructure—electricity lines and land—bets on a surge in demand for computational power driven by AI.

The arbitrage object has changed, but the logic of timing arbitrage remains the same. What Wu Jihan is truly buying is a position: “Whoever wins in the AI era will have to pay my electricity.” Don’t try to predict which technology path will dominate; just control the energy entry points for all of them.

Amazon doesn’t guess which internet company will succeed—they just rent servers to everyone. AT&T doesn’t care what you say on the phone—what matters is whether traffic flows. Industry evolution always follows the same pattern: from selling products, to selling services, to collecting infrastructure rents.

Wu Jihan is buying this execution window with billions of dollars. He is waiting for the wave of AI demand to catch up with the ever-increasing debt cycle. In two to three years, the answer will be revealed: whether this strategy has transformed Bitdeer into a global infrastructure player or is just a costly financial adventure ending in a hard lesson.