I just saw some news about Bitdeer that caught my attention. Last week, they cleared all their Bitcoin holdings—zero balance. They previously still had 943 BTC, but everything was sold. This is actually a pretty significant signal if there's a change in their fundamental strategy.

Here's the story. Bitdeer is one of the largest Bitcoin mining companies in the world, with a hashrate of 63.2 EH/s or about 6% of the total Bitcoin network. But now Wu Jihan, its founder, is making an aggressive pivot toward AI infrastructure. They’re no longer just focusing on mining, but buying and developing data centers for cloud computing.

For this expansion, their debt has already reached $1.3 billion. In less than two years, they’ve gone through several funding rounds: Tether invested $100 million in 2024, then their first convertible loan of $150 million, the second of $360 million, followed by a combined $400 million in bonds and equity in November 2025, and most recently in February 2026, another $325 million. All the money is flowing into land acquisitions and data center construction.

Their global electricity capacity now stands at 3,002 MW, with 1,658 MW already operational. For comparison, Google and Microsoft data centers usually range from 100 to 300 MW. So Bitdeer is basically assembling infrastructure equivalent to 10-30 massive data centers within a single company.

Three main locations: Rockdale, Texas, 563 MW ( focused on mining, stable cash flow), Clarington, Ohio, 570 MW ( core to their AI strategy, but currently in litigation—local steel plant suing them), and Tydal, Norway, 175 MW ( hydro power, low cost, lowest risk, scheduled to finish by late 2026).

Now, here’s the interesting part. They’re also developing their own mining chips—SEALMINER series. SEAL03 is already competitive in the market, SEAL04 aims for 5 joules per terahash, which could outperform all competitors. Margins from chip development are over 40%, much higher than mining itself. This is basically a repeat of what Wu Jihan did with Bitmain.

But there’s a problem. Revenue from AI/HPC cloud in 2025 is still very small—less than 2% of total revenue. Their GPU count increased drastically from 584 to 1,792 in three months, but utilization dropped to 41% because the new machines aren’t fully productive yet. Their annual debt service is over $650 million assuming a 5% interest rate, while revenue isn’t enough to cover that.

Analysts estimate that if HPC capacity is fully realized, potential annual revenue could be between $850 million and $2 billion. But all of this depends on: construction finishing on time, securing long-term contracts from hyperscalers, and GPUs running at full capacity. Currently, none of these conditions are met.

The most urgent risk right now is the Clarington, Ohio lawsuit. If this project is delayed by two years, the entire timeline collapses. This is literally their biggest single point of failure—not debt, not stock price, but a steel plant suing them.

Their debt structure is actually quite cleverly designed with maturities spread across 2029, 2031, and 2032. This acts as a buffer for renegotiation. If their timeline proceeds smoothly: Tydal operational by late 2026, Clarington begins construction in 2027, both assets fully operational by 2028-2029, revenue reaching the billion-dollar level, and by 2029, when the first bonds mature, conversion to equity might be an option instead of cash repayment.

But here’s the thing—markets are skeptical. Keefe Bruyette downgraded their target from $26.50 to $14, current price around $8. Wall Street is basically saying: show us the revenue first before we believe this transformation story.

So what is Wu Jihan doing by selling all his Bitcoin? He’s basically doing time arbitrage again, just like before with mining. Back then, he bet on Bitcoin’s price increase; now he’s betting on a demand explosion for compute power in the AI era. He’s not guessing who will win, he’s just controlling the entry point—electricity. Like Amazon with AWS or AT&T with infrastructure. Whoever wins the AI race will still have to pay electricity to Bitdeer.

So in the next 24 hours or in the near term, what to watch is the litigation development in Clarington and whether their construction timeline can stay on track. This is no longer about Bitcoin price or mining sentiment, but about pure infrastructure execution. The next two to three years will determine whether Wu Jihan’s strategy is genius or an expensive gamble.