Table of Contents:
70 Billion Dollars in AI Contracts: Why Bitcoin Miners are Now Building Data Centers
In a significant shift within the AI infrastructure market, Bitcoin miners are transitioning to hosting services, capitalizing on their existing resources. An estimated $70 billion in cumulative AI and high-performance computing (HPC) contracts have already been announced within the publicly traded mining sector. This trend is driven by the realization that those who already possess power connections, space, and cooling capacities can rent these out for AI workloads, often more profitably than mining Bitcoin.
"Mining stocks rose over 50% in 2026, while Bitcoin itself fell by 17%," highlighting the changing dynamics in the sector.
Companies like Hut 8 have secured a 15-year lease worth $9.8 billion for the 352-MW Beacon Point campus, while TeraWulf has signed over $12.8 billion in long-term contracts by the end of 2025. Notably, IREN entered into a five-year AI cloud contract with Microsoft valued at $9.7 billion, aiming for an annual revenue of $3.1 billion. Core Scientific is also converting 300 MW in Pecos, Texas, for AI use, planning to scale up to 1.5 GW, financed in part by selling 1,992 Bitcoin worth $175 million in March 2026.
Get $500 free Bitcoin mining for a free testing phase:
- Real daily rewards
- 1 full month of testing
- No strings attached
If you choose to buy after testing, you can keep your mining rewards and receive up to 20% bonus on top.
By the end of the year, CoinShares estimates that up to 70% of revenues from listed miners could come from AI. Core Scientific already sees 39% of its revenue from AI colocation, while TeraWulf reports 27%. This represents a fundamental reevaluation of the business model, although the transition requires significant capital investment, and not every miner will successfully navigate this transformation.
Energy as a Strategic Asset Beyond Data Centers
At the heart of these developments lies a common bottleneck: energy supply. Those who can provide power and thermal capacity will protect their margins, as the competition shifts from pure chip design to physical infrastructure. The capital-intensive nature of the underlying compute infrastructure is exemplified by Roche's launch of the largest hybrid-cloud AI factory in the pharmaceutical industry, featuring over 3,500 NVIDIA Blackwell GPUs in the US and Europe, with an estimated investment of $300 to $500 million.
This investment has already yielded results, with an oncology molecule developed 25% faster and a backup molecule created in seven months instead of over two years. With R&D expenditures of $14 billion in 2025, this investment is expected to pay off quickly, as AI compute becomes a strategic asset across various industries, further driving energy demand.
4.2% Inflation: How the Interest Rate Environment Could Slow the CapEx Boom
Despite the structural growth potential, investors should not overlook the macroeconomic environment. The Federal Reserve held the interest rate steady at 3.50 to 3.75% for the fourth consecutive time on June 18, due to US inflation rising to a three-year high of 4.2% year-over-year in May. Energy prices surged by 23% amid the Iran conflict, with gasoline prices increasing by 40% to an average of $4.60 per gallon. Following the Fed's decision, Treasury yields rose, and stock indices declined.
For investors in AI infrastructure, this has direct implications: persistently high interest rates increase the cost of capital-intensive investments that drive the boom. Those betting on operational beneficiaries should keep an eye on the financing costs within the industry.
What This Means for Investors
The infrastructure trend surrounding power and cooling is real and supported by solid contractual figures. From cooling and power specialist Vertiv to AI renters emerging from the Bitcoin sector, the bottleneck is shifting from computing power to energy. Investors should clearly differentiate between operational beneficiaries with real revenues, like Vertiv, and business models undergoing transformation, such as miners. Additionally, it is crucial to remember that in a hot CapEx cycle, coupled with a tight interest rate environment, the risk of over-investment also rises. The key question is not whether AI infrastructure will be built, but who can finance it profitably.
In summary, the landscape of AI infrastructure is rapidly evolving, with Bitcoin miners pivoting towards more lucrative opportunities in AI hosting, driven by their existing capabilities. However, the macroeconomic environment poses challenges that investors must navigate carefully.
Sources:
