I just noticed something interesting about Bitdeer's strategy. Last week they posted an production update: after mining 189.8 BTC, they immediately sold it. Remaining stock is 943.1 BTC? Also sold. Their Bitcoin balance is now zero. This is no coincidence—it's a very deliberate strategic decision by Wu Jihan.

From the beginning, Bitcoin mining has actually been a game of arbitrage over time. You take electricity and machines today, exchange for Bitcoin tomorrow. This logic has played out for twelve years. But now Wu Jihan is changing the target of this game. It’s no longer about coin prices—now about the long-term value of computational power demand in the AI era. The method is changing: from electricity consumption to earn coins, to borrowing huge amounts of money to buy land.

In the same week as clearing Bitcoin, Bitdeer completed a new debt issuance of $325 million. Their total debt now reaches $1.3 billion. The debt is real, the land purchases are real, but the results of this fierce battle will only be revealed in 2029.

Bitdeer was founded in 2018, originally as a mining machine sharing platform. Now it’s one of the largest listed mining companies in the world, with its own mining computational power of 63.2 EH/s—highest among listed mining companies worldwide, about 6% of the total Bitcoin network hash rate. But Wu Jihan no longer wants to sell computing power. He wants to control electricity.

By early 2026, Bitdeer’s total global power capacity reaches 3,002 MW, with 1,658 MW operational and 1,344 MW under construction or pending. For comparison, large hyperscale data centers from Microsoft and Google are usually 100 to 300 MW. So 3,002 MW is equivalent to combining the electricity needs of 10 to 30 of Google’s large data centers within a single company. Their capacity is theoretically very impressive.

This $1.3 billion debt is mainly to secure assets of global electricity land, preparing for the transition to AI data centers. First, Rockdale, Texas—563 MW already operational, focused on mining with stable cash flow. Second, Clarington, Ohio—570 MW, 30-year lease, electricity contracts signed, planned to complete by Q2 2027, positioned as HPC/AI center. This is the core of the entire AI transition plan. Third, Tydal, Norway—175 MW, transforming from a mine into an AI data center, scheduled for late 2026, with an effective IT capacity of 164 MW. Hydroelectric sources, competitive energy costs. The transformation is much cheaper than building new.

Land, electricity, and server space—these three assets are called the 'most difficult to imitate' in the AI industry. Bitdeer has accumulated all of them over ten years of mining operations. There’s one rarely discussed but noteworthy point: SEALMINER. Bitdeer is not only building mining centers but also developing its own mining chips. The SEAL series is now in its third generation, with SEAL03 energy efficiency at 9.7 joules per terahash. The A3 Pro, set for mass production in September 2025, is already among the top globally. SEAL04 aims for 5 joules per terahash—if achieved, it will surpass all mass-market mining machines. The gross margin of these chips is over 40%, far higher than mining itself. This echoes what Wu Jihan did at Bitmain: from buying other people’s shovels to making their own.

Now about the debt. In less than two years, several funding rounds occurred. May 2024, Tether invested $100 million, becoming the second-largest shareholder. Three months later, the first convertible loan of $150 million at 8.5% annual interest. November 2024, the second convertible loan of $360 million, with interest reduced to 5.25%. November 2025, a combined offering: $400 million in convertible bonds and 148.4 million additional shares. February 2026, again with $325 million in convertible bonds and 43.5 million additional shares, plus repurchasing $135 million of old bonds maturing in 2029, extending to 2032. Total exceeds $1.4 billion.

Each time bonds are issued, Bitdeer’s stock price drops 10 to 17%. This has become a market reflex. Fortunately, the company always manages to raise funds. The core of the debt structure is convertible bonds. The new bonds due 2032 have an initial conversion price of about $9.93, a 25% premium over the simultaneous offering price of $7.94. When the stock price reaches that level, bondholders will convert to shares instead of cash. The company essentially doesn’t need to repay the debt as long as the stock price rises. The logic of convertible bonds is to seek a rise in the stock’s own price—basically a bet on whether the AI narrative will be accepted by the market.

The annual interest burden, assuming an average rate of 5% and principal of $1.3 billion, results in over $650 million in interest expenses per year. Meanwhile, AI/HPC cloud revenue throughout 2025 will not reach one-tenth of six-month interest payments. Currently, this interest is entirely dependent on continuous debt issuance to keep turning over. The pressure is immense.

But what is the potential profit from AI? The AI business now generates $10 million per year, less than 2% of total revenue. For a company with a market cap approaching $2 billion, this figure is almost negligible. Bitdeer’s GPU count increased from 584 to 1,792 in three months, tripling. But utilization rate dropped from 87% to 41%, mainly because the machine deployment was too rapid. B200/GB200 are still in customer testing, not yet generating revenue. Electricity is in place, machines are being installed, the denominator is jumping, but revenue has not yet caught up.

Roth/MKM estimate that fully realized HPC capacity could generate annual revenue of $850 million. Management is more aggressive: all 200 MW allocated to AI cloud, annual revenue could exceed $2 billion, three times the 2025 mining revenue. But both figures depend on three conditions: project completion on time, securing long-term contracts with hyperscalers, and GPUs running at full capacity. None of these are currently met.

This is the war Bitdeer is playing: mining to support AI, AI shaping the dream, but whether the dream can come true still depends on execution over the next two to three years.

Bitdeer’s debt structure is actually designed to be more stable than it appears. High-leverage companies usually fail for the same reason: maturing debt concentrates, insufficient cash, forced asset sales. Bitdeer has set maturity dates for three series of convertible bonds—2029, 2031, and 2032—considered as a deliberately created buffer zone. The first batch matures, Tydal and Clarington are theoretically in place. The second batch matures, AI revenue should speak for itself. The third batch matures, the market will evaluate what this company really is.

But convertible bonds buy time, and Wall Street isn’t buying it. Keefe Bruyette lowered its target price from $26.50 to $14. The current stock price is around $8. The market signals are very realistic: the transformation story must show revenue. But all this pressure gives Wu Jihan the most needed—and most brutal—thing: time.

The smoothest possible scenario might look like this: late 2026, Tydal renovation completed, the 164 MW hydroelectric data center in Norway begins operation, European customer contracts signed. 2027, Clarington wins in court, the 570 MW project in Ohio officially starts construction, major US customers follow. 2028 to 2029, two core assets fully operational, revenue reaching billions, analysts rebrand Bitdeer from a discounted mining company to a premium AI infrastructure firm. 2029, the first debt matures, debt holders see the stock price and most likely choose to convert rather than cash out.

In every fierce battle, Wu Jihan must be on time. Clarington remains. In the same Ohio industrial park, there’s a steel producer called American Heavy Plate Solutions that signed a 30-year lease for 9.9 acres in 2018. They sued Bitdeer: building an AI data center would disrupt electricity, roads, railways, and shared communication lines, violating restrictions. Their demand: a court order for a permanent injunction to prevent Bitdeer from starting construction. Clarington accounts for 42% of the pipeline under construction. If delayed, the entire schedule must be rewritten. So the biggest single-point risk for Bitdeer right now isn’t debt, isn’t stock price—it’s a steel plant.

In mining, there’s also no time to rest. February 2026, Bitcoin network difficulty jumps 14.7%, the largest single increase since May 2021. With the same electricity, fewer coins are mined. Gross margin in Q4 drops from 7.4% a year ago to 4.7%. This mining branch is gradually becoming thinner.

The worst-case scenario is clear: a prolonged two-year lawsuit with Clarington, halted construction. Tydal faces delays, GPU utilization remains around 41%. The first debt matures in 2029, cash on the balance sheet isn’t enough, refinancing is forced, shares continue to dilute, and the conversion threshold becomes even harder to reach. Two paths, both very real.

In the mining world, there’s a tradition: mining coins is a belief, a support for Bitcoin’s long-term value. MARA mines 53,250 BTC, Riot mines 18,000, Strategy mines 710,000. The more mined, the more the market believes you’re confident. Bitdeer now holds zero. The official explanation: selling Bitcoin to provide liquidity for land purchases. This isn’t wrong. Industry peers are moving the same way—Riot sold $200 million worth of Bitcoin for AI expansion, Bitfarms is divesting as a Bitcoin company, MARA is building HPC. But there’s something more fundamental here.

Since day one, the mining industry has always bet on one thing: that something in the future will be more expensive than the current cost. Ten years ago, mining bet on rising coin prices. Now, land purchases bet on a surge in demand for computational power. The object has changed, but the arbitrage logic over time has never changed. What Wu Jihan is truly buying is the position that whoever wins must pay for electricity. Don’t guess the path, just control the entry point to that path.

Amazon doesn’t guess which internet company will win, it just rents servers to everyone. AT&T doesn’t care what you say on the phone, only whether you call or not. From selling products, to selling services, to leasing—industry evolution always follows one path. The only question is whether you actively move closer or are pushed by others. Anything that produces energy is called a critical asset in this era, and Wu Jihan is buying this window with billions of dollars. He’s waiting for money from AI to chase his debt speed.