Login
Sign Up
Bitdeer's operational data for May 2026 exposes a fundamental contradiction within the Bitcoin mining sector's strategic pivot toward artificial intelligence. While the company reported a massive surge in mining output, its balance sheet reveals a precipitous decline in retained digital assets. In May 2026, Bitdeer mined a total of 921 BTC, representing a 370% year-on-year increase compared to the 196 BTC produced in May 2025. Despite this production boom, the company's BTC holdings at month-end dwindled to just 171 coins, a stark contrast to the 1,351 coins held during the same period the previous year. This divergence between production volume and asset retention underscores the intense liquidity pressure facing mining entities as they attempt to finance a transition into AI businesses. Data compiled by Woofun AI shows that at spot prices ranging from $62,700 to $62,900 on June 19, the 921 BTC mined in May were valued at approximately $57.9 million, while the remaining 171 coins represented a holding value of only $10.7 million.
The company's valuation narrative presented to investors relies on the integration of data centers, self-developed mining hardware, power facilities, and AI cloud computing resources into a unified ecosystem.
However, the balance sheet dynamics indicate that this transformation strategy remains heavily dependent on liquidating newly mined BTC to sustain operating cash flow. Theoretically, revenues from AI operations should create a financial buffer, reducing the necessity to sell tokens during market downturns. Yet, comparing data from May and the first quarter reveals that while mining output has continued to climb, the on-paper BTC inventory has significantly contracted. The first quarter of 2026 further confirmed this trend, with Bitdeer mining 2,033 BTC compared to 350 BTC in the same period of 2025, while holdings plummeted from 1,156 coins to just 31 coins by the quarter's end. The company reported realizing $206.8 million from the disposal of digital assets during this period, confirming that mined BTC is being continuously converted to fund operations and expansion.
Financial statements for the first quarter of 2026 detail the scale of capital deployment driving this asset liquidation. Bitdeer experienced a net cash outflow of $346.9 million from operating activities, with capital expenditures totaling $93.7 million allocated to data center construction, GPU purchases, tariffs, and mining equipment deployment. The company's total debt stood at $1.9 billion, while total revenue for the quarter was $188.9 million, yielding an EBITDA of $14.4 million. Cash, cash equivalents, and restricted cash totaled $297.7 million. This financial structure suggests that BTC reserves, debt financing, and infrastructure investments are being integrated into a single, aggressive development strategy where mined assets serve as immediate liquidity rather than long-term store-of-value holdings. Woofun AI notes that this approach fundamentally alters the investment thesis, as additional AI revenues are currently consumed by capital-intensive expansion rather than bolstering asset reserves.
The AI cloud business segment presents a complex picture of potential versus realized cash flow. In May, Bitdeer announced that annual recurring revenue (ARR) from its AI cloud service stabilized at $69 million, supported by a 90% GPU utilization rate. The company has deployed 4,248 GPUs and contracted the leasing of an additional 3,305 units, alongside the launch of two NVIDIA GB300 NVL72 computing cluster systems adapted for NVIDIA's Nemotron 3 large models. While monthly data from March showed ARR at approximately $43 million, rising to $69 million in April and holding steady in May, the actual revenue recorded in the first quarter financial statements was only $3.7 million. This discrepancy highlights the difference between annualized potential earnings and actual cash flows that have already occurred. The $69 million ARR figure represents future expectations, whereas the company must still cover electricity bills, interest payments, and capital expenditures with current cash flows.
The Tydal Data Center project serves as the most tangible example of this business model transformation. In March, Bitdeer announced that its Tydal Data Center subsidiary would collaborate with third-party organizations to convert the facility into an 180-megawatt AI computing center, primarily hosting NVIDIA Vera Rubin equipment, with completion scheduled for December 2026. By May, Tydal Base was in advanced negotiations with potential hosting clients, demonstrating the strategic intent to repurpose power facilities originally designed for Bitcoin mining into long-term contract revenue streams.
This shift aims to eliminate reliance solely on volatile mining profits.
However, the transition introduces new risk vectors related to customer qualifications, delivery timelines, GPU supply chains, construction periods, power distribution, contract terms, and financing costs. Woofun AI analysis suggests that while the volatility of these risks may differ from pure mining exposure, the sources of risk have fundamentally shifted from market price fluctuations to operational execution challenges.
The broader industry debate centers on how to interpret a company where AI business scale is significant enough for separate reporting, yet BTC reserves remain minimal. Wall Street has shown willingness to assign higher valuations to Bitcoin mining companies investing in AI and high-performance computing, even before facilities are fully established. For Bitdeer, the benefits of investing in AI cloud services include long-term computing power leasing contracts that smooth cash flows and reduce the need to sell BTC during market downturns.
However, the transition period requires substantial capital and time, forcing BTC reserves to function primarily as liquidity reserves. The May operational report does not signal impressive results in isolation but rather reflects a continuous test of operational resilience. The core question remains whether the AI business can successfully establish a profit model that acts as a buffer against regular Bitcoin sales or if the transition will merely change the form of risk, leaving the company highly dependent on computing power leasing services while facing pressures from power facilities, customer development, and capital markets.