I recently saw something interesting about Bitdeer that indicates a major shift in the Bitcoin mining industry. Wu Jihan's company is now moving beyond just mining into the AI data center business. This is not just a pivot — it's a complete overhaul of the entire strategy.

Last week, Bitdeer sold all of its Bitcoin. From 189.8 BTC weekly production to stock, everything went to the market. Zero Bitcoin holdings. This sends a different signal among mining companies. MARA has 53,250 BTC, Riot has 18,000 BTC, but Bitdeer now has zero. The government explanation is straightforward — liquidity is needed to buy land. That’s true, but a deeper strategy is hidden behind it.

In fact, Bitdeer has raised $13 billion in debt. Before issuing a new bond of $325 million in February, it already had $10 billion in debt due by the end of 2025. Multiple rounds over two years — $1 billion investment from Tether, then a $150 million convertible bond, then $360 million, then $400 million, and now $325 million. Total over $14 billion.

What is this money being used for? Three main areas — 563 MW of mining in Rockdale, Texas, 570 MW of AI data centers in Clarington, Ohio, and 175 MW of hydro power in Tidal, Norway, converting into AI. These are the three 'hardest copyable things' — land, electricity, and data center infrastructure.

I see an interesting detail here — Bitdeer is also manufacturing its own mining chips. The SEAL series has reached the third generation, with SEAL03 achieving 9.7 joules per terahash energy efficiency. SEAL04 is heading toward 5 joules. This could impact the entire Bitcoin mining machine market because better efficiency means a big competitive advantage. Chips have profit margins over 40% — much more than mining.

But there is pressure too. In Q4, gross margin fell from 7.4% to 4.7% because Bitcoin network difficulty jumped 14.7% in February — the largest single increase since May 2021. Fewer coins are being mined with the same electricity. The mining wheel is slowly thinning out.

On the AI side, GPU count increased from 584 to 1,792 in three months — three times. But utilization rate dropped from 87% to 41% because machines are being deployed very rapidly. B200/GB200 are currently in customer testing. Power is connected, machines are being installed, everything is growing fast — only revenue is lagging behind.

Analysts say that at full capacity, annual revenue could reach $8.5 billion. Management is more aggressive — deploying 200 MW fully into AI could be worth over $20 billion. But three conditions are set: construction must finish on time, hyper-scaler level long-term contracts must be secured, and GPUs must be fully utilized. So far, none have been fully achieved.

The biggest single-point risk? A steel factory. In Ohio, American Heavy Plate Solutions has sued, claiming Bitdeer’s AI data center sharing power, roads, railways, and communication lines will disrupt them. Clarington’s 42% construction pipeline could be stuck. If this drags on for two years, the entire timeline shifts.

Looking at the debt structure, Bitdeer has set the maturity of convertible bonds for 2029, 2031, and 2032. This is a deliberately designed buffer zone. By 2029, both Tidal and Clarington could be operational, by 2031 AI revenue could start flowing, and by 2032, the company’s true form will emerge. Three checkpoints, three chances to renegotiate.

But Wall Street didn’t give this time. Keefe Bruyette lowered its target from $26.50 to $14. The stock is currently around $8. The market signal is clear — the story of transformation will be judged solely by revenue.

In fact, Wu Jihan has bought himself a situation where ‘whoever wins, they will have to pay my electricity bill.’ Don’t bet on competition — just control the entry point. Amazon doesn’t bet on an internet company; it rents servers to everyone. This is a major strategy shift from mining to AI. Now, it remains to be seen whether this biscuit will become real in the next 2-3 years.