Bernstein initiates coverage on bitcoin mining firms TeraWulf and Cipher Digital with aggressive price targets as the sector pivots toward a new role supplying power and infrastructure for artificial intelligence (AI) data centers.
Bernstein sets aggressive targets as miners morph into AI power landlords
Bernstein has initiated coverage on bitcoin mining firms TeraWulf and Cipher Digital, setting price targets of $36 and $32 respectively, as it forecasts a rapid shift in the sector toward supplying power and infrastructure for AI data centers.
The firm expects AI-related revenue for the miners it tracks to surge from $1.2 billion in 2026 to $10.7 billion by 2030. Bernstein estimates this group will represent about 10% of all U.S. data centers currently under construction, underscoring how critical their grid-connected power portfolios have become to the AI buildout.
Analysts led by Gautam Chhugani described TeraWulf and Cipher as major holders of grid-connected power capacity, calling it the key bottleneck in building new AI computing sites. Across the industry, U.S. miners have signed 17 energy capacity agreements over the past two years, valued at more than $110 billion and covering 6 gigawatts of power, according to the report.
TeraWulf: power-led AI strategy with repurposed industrial sites
Bernstein projects TeraWulf’s AI revenue to jump from $14 million in 2025 to $1.7 billion in 2030, implying a compound annual growth rate of about 163%. EBITDA margins are forecast to reach 84%, with EBITDA rising from $106 million in 2026 to $1.4 billion by 2030.
TeraWulf has contracted 643 megawatts of capacity with Fluidstack and Core42, worth roughly $13 billion across agreements of up to 25 years. About 91% of this exposure is tied to Fluidstack, which Bernstein identifies as a key concentration risk.
The company redevelops former industrial assets, including a repurposed coal plant and an aluminum smelter. Bernstein estimates this strategy cuts buildout costs to $7.2 million per IT megawatt, compared with a typical peer range of $11 million to $13 million. TeraWulf’s 3.8-gigawatt power portfolio spans four U.S. states, which the firm says reduces regulatory risk tied to any single jurisdiction.
Bernstein’s valuation model applies a 21-times one-year forward EV/EBITDA multiple on steady-state 2030 results to arrive at its $36 price target.
The shift is already reshaping market perception. TeraWulf and other miners that have aggressively embedded themselves in the AI supply chain, such as IREN, have seen share price gains of more than 700% over the past year. TeraWulf’s recent purchase of a Kentucky campus with potential for more than 1 gigawatt of data center development further extends its infrastructure platform.
Cipher Digital: long-term AI contracts and triple-net leases
Cipher Digital’s AI revenue is forecast to rise from $19 million in 2026 to $1.2 billion by 2030, with margins stabilizing near 93%, according to Bernstein.
The company’s contracts span 495 megawatts with an estimated order value of about $11.4 billion. Roughly two-thirds of this is backed by investment-grade counterparties, including major cloud service providers. Around 48% of contracted revenue is linked to AWS, creating a significant customer concentration that Bernstein flags as a core operational risk.
Cipher has shifted newer deals toward triple-net lease structures, which push operating costs to the tenant and are designed to boost returns and reduce financing costs. Its AWS agreement helped Cipher secure $2 billion in project funding at a 6.125% interest rate, about one percentage point below earlier financings.
This financial structure has drawn attention beyond Bernstein. Analysts at Jefferies and H.C. Wainwright have raised their targets on Cipher as the company pivots more fully into high-performance computing infrastructure.
IREN tops Bernstein list as sector re-rates on power, not hashrate
Bernstein named IREN as its top pick in the group, assigning a $100 target based on its 5-gigawatt global power portfolio and its positioning as a large-scale AI infrastructure provider.
Other covered names include Core Scientific with a $32 target, Riot Platforms at $30, and CleanSpark at $24. Marathon Digital was maintained at market-perform with a $17 target, reflecting a more cautious view as peers accelerate their transition to AI-focused infrastructure.
The shift in analyst focus mirrors a broader market re-rating. Rather than valuing miners primarily on bitcoin hashrate, traders are increasingly focusing on the size and quality of a company’s power pipeline and the durability of its long-term high-performance computing contracts.
Core Scientific has leaned into this narrative, recently outlining plans to expand its Oklahoma campus to roughly 1.5 gigawatts. Its shares reached an all-time high in early June 2026, reflecting optimism about its infrastructure strategy. CleanSpark, meanwhile, has faced tougher questions after recent earnings disappointed, despite management’s emphasis on evolving into a data center development platform.
AI power crunch sets backdrop for miners’ new role
Bernstein’s analysis unfolds against what it calls a structural power shortage for AI data centers. Goldman Sachs Research expects U.S. data center power consumption to more than double between 2025 and 2027, eventually consuming about 8.5% of national peak power.
Some analysts now predict that AI data center development could be constrained by 2027 as local grid capacity is exhausted. That dynamic has turned securing grid-connected power into the decisive factor for AI infrastructure, rather than chip supply alone.
Digital asset miners are positioned to benefit because they have already negotiated large-scale power deals and built out industrial sites capable of hosting energy-intensive computing. The 6 gigawatts of contracted capacity signed by U.S. miners over the past two years represent a meaningful slice of the power needed for new data centers under construction, Bernstein noted, accelerating their reclassification from bitcoin producers to high-performance computing landlords.
Concentration, grid approvals, and financing remain key risks
Bernstein highlighted several risks that could undermine the bullish AI thesis for miners.
First, revenue concentration is significant. Fluidstack accounts for about 91% of TeraWulf’s contracted revenue, while AWS makes up about 48% of Cipher’s. Any shift in these customers’ strategies, renegotiations, or non-renewals could materially affect earnings and valuations.
Second, ongoing approvals for grid connections present timing and execution risk. Regulatory reviews, permitting delays, and local opposition could slow or cap the ability of miners to bring new capacity online at the pace implied by current AI demand forecasts.
Third, the capital intensity of this buildout is substantial. Large-scale data center campuses require multi-billion-dollar funding, and while recent deals have shown improving financing terms, balance sheet pressure remains high. Bernstein noted that several miners reported net losses in their first-quarter 2026 results, underscoring the heavy upfront investment required before AI revenues scale.
For traders, Bernstein’s coverage signals that bitcoin mining equities are being recast as power and infrastructure plays tied to the AI cycle, with outsized upside tied to execution — and equally elevated risks around concentration, permitting, and capital access.
Want deeper insight into mining economics? Explore our guide on how bitcoin mining really works and its evolving profit drivers.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
