TeraWulf's Rebound Depends on Turning Power Into Contracted AI Revenue

TeraWulf shares rose about 6% as AI infrastructure stocks rebounded from a sharp selloff, but the durable investment case rests on a more important transition: moving from Bitcoin mining to contracted high-performance-computing revenue.

The company's official Q2 2026 presentation reported $44.8 million of total revenue and $31.9 million of gross HPC lease revenue, more than 70% of the total. That is a meaningful change from a year earlier, when the segment contributed no revenue. TeraWulf also reported $3.0 billion of cash, cash equivalents and restricted cash.

Liquidity does not remove financing risk. The same presentation calculated total debt at $5.7 billion and net debt at $2.7 billion. New campuses require billions of dollars before rent begins, and some projects will not deliver capacity until the second half of 2027 or later.

Power access is TeraWulf's main strategic asset. At Lake Mariner, the company reported 102 megawatts of revenue-generating critical IT capacity and another 336 megawatts under construction. It also described a contract-first development model intended to match construction with long-duration customers.

The remaining risk is tenant concentration and timing. A project without a committed, creditworthy tenant can consume capital while producing no lease income. Schedule changes can also push cash receipts further out even when a contract remains in place. Investors should distinguish contracted capacity, capacity under construction and speculative development pipeline.

The rebound in TeraWulf, Applied Digital and IREN shows that financing fears can reverse quickly. It does not prove that the economics of every project have improved. The better evidence will be new lease signings, delivered megawatts, customer credit support and cash rent.

TeraWulf has made measurable progress toward an AI-hosting model. The next phase is harder: repeat the conversion while keeping debt, dilution and construction schedules under control. That execution, not the daily direction of AI infrastructure shares, will determine whether the transition creates shareholder value.