Overview
AI infrastructure stocks are rallying together as investors broaden the artificial intelligence trade beyond GPUs and semiconductor suppliers into the physical infrastructure required to turn chips into usable computing capacity.
On August 12,
Reuters reported that CoreWeave shares rose more than 19%, Nebius Group gained about 23% at the time of the report, IREN advanced roughly 8%, and Applied Digital climbed about 4%. Nebius extended its rally later in the session as investors digested its own stronger-than-expected quarterly results.
The common catalyst was evidence that AI computing demand remains ahead of available capacity.
CoreWeave reported $2.575 billion of Q2 revenue and approximately $104 billion of revenue backlog while raising its 2026 capital expenditure outlook to $35 billion to $39 billion. Nebius then reported 454% revenue growth and 514% growth in its core AI cloud business.
IREN and Applied Digital occupy different parts of the infrastructure chain, but both benefit from the same underlying constraint.
IREN is converting a large portfolio of power and data center assets previously associated with Bitcoin mining into vertically integrated AI cloud infrastructure. Applied Digital is increasingly focused on developing and leasing high-density data centers to customers including CoreWeave.
The rally therefore signals a broader shift in the AI investment narrative. Investors are moving beyond the question of who sells the leading GPU and toward a second question: who controls the power, data centers, cloud platforms and deployment capacity needed to operate those GPUs at scale?
Key Takeaways
CoreWeave, Nebius, IREN and Applied Digital rallied together on August 12 as AI infrastructure earnings strengthened confidence in the data center buildout.
CoreWeave reported $2.575 billion of Q2 revenue and approximately $104 billion of revenue backlog, excluding more than $25 billion of additional commitments secured in early Q3.
CoreWeave raised its 2026 CapEx outlook to $35 billion to $39 billion as near-term compute capacity remained effectively sold out.
Nebius generated $582.3 million of Q2 revenue, up 454% year over year, while AI cloud revenue increased 514% to approximately $574.9 million.
Nebius signed four major AI cloud contracts during Q2 with average total contract values above $1 billion each.
IREN raised its year-end 2026 AI cloud annualized run-rate revenue target to more than $4 billion, with approximately 85% already under contract.
IREN secured another $2.8 billion of multi-year AI cloud contracts in July as its transition from Bitcoin mining toward AI infrastructure accelerated.
Applied Digital is positioned differently from the cloud operators. Its business increasingly focuses on developing and leasing high-density data center capacity, including 400 MW contracted to CoreWeave at Polaris Forge 1.
The stocks can respond to the same infrastructure cycle while still carrying very different financing, execution, customer and valuation risks.
CoreWeave Earnings Became the Catalyst for the Sector
CoreWeave's Q2 report addressed one of the biggest concerns surrounding the AI infrastructure cycle.
The industry is building an enormous amount of capacity, but is there enough underlying demand to absorb it?
The latest numbers suggest that demand remains strong.
According to
CoreWeave's Q2 2026 results, revenue reached $2.575 billion compared with $1.212 billion a year earlier, an increase of approximately 112%.
Revenue backlog stood at approximately $104 billion as of June 30.
That figure excluded more than $25 billion of additional net customer commitments secured in early Q3.
Reuters reported the backlog at a more precise $104.2 billion, up from $99.4 billion three months earlier.
CoreWeave management also said near-term capacity was effectively sold out, allowing the company to sign new compute contracts on increasingly favorable commercial terms.
That is a crucial signal for the wider AI infrastructure sector.
Higher CapEx Is Being Read as Evidence of Demand
Reuters reported that CoreWeave raised its 2026 capital expenditure forecast to $35 billion to $39 billion from a previous range of $31 billion to $35 billion.
Higher spending would normally create concerns about free cash flow and financing requirements.
The interpretation changes when existing capacity is already heavily committed.
CoreWeave needs additional data centers, power, GPUs and networking infrastructure to convert contracted demand into revenue-generating compute capacity.
As long as new capacity can be contracted at attractive rates, incremental capital spending can support future revenue rather than simply create excess assets.
That explains why investors initially treated the higher CapEx forecast as a sign of continuing AI demand rather than an immediate negative.
CoreWeave Demand Flows Through the Supply Chain
CoreWeave does not produce every component required to operate its AI cloud.
It buys accelerated computing systems, deploys servers and networking equipment, consumes large amounts of power and leases or develops data center capacity.
The economic effects of each new CoreWeave compute contract therefore extend well beyond CRWV itself.
Chip and memory suppliers benefit from hardware demand.
Data center developers can benefit from additional space and power requirements.
Other specialized cloud providers can benefit if strong utilization and contract pricing demonstrate that AI compute remains scarce.
This explains why a CoreWeave earnings report can move CRWV, NBIS, IREN and APLD at the same time even though their businesses are structurally different.
Nebius Provided a Second Confirmation of AI Cloud Demand
If CoreWeave had been the only specialized AI cloud company reporting strong results, investors could have attributed the numbers to market share gains or customer-specific factors.
Nebius made the demand signal broader.
Nebius AI cloud revenue reached $574.9 million, up 514% year over year and representing approximately 98% of total group revenue.
Annualized run-rate revenue for the AI cloud business reached $3 billion at the end of June, up 598% from a year earlier.
These are not peripheral business gains.
They show that the core AI compute platform is scaling rapidly.
Customers Are Locking In Capacity Earlier
Nebius described Q2 as its strongest commercial quarter to date.
The company signed four major AI cloud contracts during the quarter with average total contract values above $1 billion each.
Reuters reported that Nebius now has more than $40 billion in customer commitments and expects more than $9 billion of customer prepayments during 2026.
Prepayments matter because they can partially finance the infrastructure required to serve those customers.
They also suggest customers are willing to commit capital in advance to secure future computing capacity.
Nebius further disclosed that new contracts in its core AI cloud business averaged annual contract values above $20 million per megawatt.
That provides another signal that available AI capacity remains valuable.
Nebius Reinforced the CoreWeave Read-Through
CoreWeave and Nebius do not have identical platforms or capital structures, but both sell specialized computing services into the AI market.
When both companies report rising commitments, improving utilization and strong contract economics at approximately the same time, the probability increases that the underlying demand is industry-wide rather than company-specific.
That helped turn the August 12 move into a sector rally.
Investors were not simply rewarding one earnings beat.
They were repricing the economics of specialized AI cloud capacity.
Why IREN Joined the Rally
IREN represents a different pathway into AI infrastructure.
The company is familiar to many crypto investors because of its Bitcoin mining history and large portfolio of power-intensive data centers.
In 2026, however, the company's growth strategy has increasingly shifted toward AI cloud infrastructure.
According to
IREN's Q3 FY2026 business update, the company had already secured a five-year $3.4 billion AI cloud contract and entered into a broader strategic partnership with NVIDIA covering its global 5 GW secured data center pipeline.
IREN explicitly described its financial transition as a move from Bitcoin mining toward AI cloud.
That makes the company one of the clearest links between the public Bitcoin mining sector and the AI infrastructure trade.
New Contracts Strengthened the AI Transition
On July 20, IREN announced another $2.8 billion of multi-year cloud services contracts with leading AI developers.
Approximately 85% of that target was already under contract.
IREN's customer base includes Microsoft, NVIDIA, Perplexity, Figure AI, Together AI, Fluidstack, Fireworks AI, Fal AI and Hume AI.
The company plans to deliver approximately 480 MW of AI cloud capacity by the end of 2026 and targets roughly 1.2 GW in 2027.
As of June 30, IREN reported approximately $7.6 billion of cash and cash equivalents, including restricted cash associated with financing for specific GPU deployments.
Power May Be IREN's Most Important Asset
IREN's Bitcoin mining history gave the company something that has become increasingly valuable in the AI cycle: access to large-scale power and data center infrastructure.
In Bitcoin mining, electricity is converted into hash rate and ultimately into mining revenue.
In AI infrastructure, the same grid connections, land and facilities can support high-value computing equipment under multi-year cloud contracts.
If an AI customer can produce a higher and more predictable economic return from a megawatt of capacity than Bitcoin mining, the value of the underlying power asset changes.
That is why strong CoreWeave and Nebius demand can lift IREN even without a new IREN earnings report on the same day.
Investors are reconsidering what IREN's existing power and development portfolio may be worth in an AI-constrained infrastructure market.
Why Applied Digital Is Part of the Same Trade
Applied Digital sits at another point in the value chain.
Its core strategy increasingly resembles an AI data center developer and infrastructure landlord rather than a company primarily buying GPUs and reselling compute by the hour.
If CoreWeave and Nebius represent demand for AI cloud capacity, Applied Digital represents part of the physical infrastructure required to supply it.
CoreWeave Is Already a Major Applied Digital Customer
The commercial link between the companies is direct.
That makes CoreWeave's latest demand data particularly relevant for APLD.
If CoreWeave's backlog continues expanding and near-term capacity remains effectively sold out, investors have more reason to believe that contracted third-party data center capacity will remain economically valuable.
CRWV and APLD therefore are not simply two stocks reacting to the same headline.
CoreWeave is an AI compute operator and tenant. Applied Digital supplies part of the physical capacity supporting that compute.
Data Center Revenue Is Now Appearing in Applied Digital's Results
Data center rental and other revenue reached $50.6 million, including $44.1 million of base rent.
Adjusted EBITDA increased to $42.4 million from approximately $1 million a year earlier.
The figures show that Applied Digital's AI data center strategy is moving from construction plans into recognized revenue.
The company is also continuing to expand its long-term lease portfolio.
Reuters reported in June that Applied Digital signed a 15-year lease worth approximately $5.2 billion with an investment-grade U.S. hyperscaler covering 210 MW of computing capacity.
The APLD rally therefore reflects more than thematic association with AI.
Investors are assigning value to long-duration data center contracts and scarce power capacity.
Why Capital Is Moving Beyond GPUs
The earliest phase of the AI investment trade concentrated heavily on semiconductor suppliers.
The logic was straightforward.
Training increasingly large models required advanced accelerators, those accelerators were scarce, and the companies supplying them captured substantial economic value.
The infrastructure cycle is now becoming more complex.
A GPU sitting in a warehouse is not usable AI compute.
It has to be installed in a server, connected to high-speed networking and memory, placed in a data center, supplied with reliable electricity and cooled at extremely high density.
Every one of those requirements can become a bottleneck.
Investors Are Pricing Physical Scarcity
CoreWeave says near-term compute capacity is effectively sold out.
Nebius is signing billion-dollar contracts and receiving substantial customer prepayments.
IREN is converting grid-connected power into AI cloud capacity.
Applied Digital is monetizing large data center campuses through long-term leases.
The businesses are different, but they all control something that cannot be solved with software alone.
That includes electricity, grid interconnections, land, high-density facilities, cooling and the operational ability to bring large computing clusters online.
As investors gain confidence that end-user compute demand remains strong, those physical assets become more valuable.
A Common Rally Does Not Mean a Common Valuation Model
The distinction is important.
CoreWeave combines rapid cloud revenue growth with extremely heavy capital expenditure and financing requirements.
Nebius operates a specialized AI cloud platform with its own customer mix, geographic strategy and capital structure.
IREN combines owned power infrastructure, data center development and AI cloud services while retaining a history and asset base linked to Bitcoin mining.
Applied Digital is increasingly a long-duration infrastructure developer and landlord, where lease economics, construction costs and financing conditions are central to valuation.
All four stocks can rise on the same industry catalyst.
They should not automatically trade at the same multiples or be analyzed using the same financial metrics.
What Could Break the AI Data Center Rally
The synchronized move supports the view that AI infrastructure demand remains strong, but it can also encourage investors to overlook the risks created by the scale of the buildout.
Capital-intensive industries are often most vulnerable not when demand is weak, but when every participant assumes strong demand will persist indefinitely.
Today's Shortage Could Become Tomorrow's Supply
CoreWeave expects up to $39 billion of capital expenditure in 2026.
Nebius is rapidly increasing computing and data center capacity.
IREN is targeting a move from hundreds of megawatts toward more than a gigawatt of AI cloud capacity.
Applied Digital continues to develop large new campuses.
If AI training and inference demand expands quickly enough, those investments can generate substantial revenue.
If capacity grows faster than customer demand, compute pricing could fall.
When pricing weakens, projects built with expensive debt or equity capital can experience a sharp decline in expected returns.
Backlog Is Not the Same as Profit
CoreWeave's approximately $104 billion backlog is substantial, but the company still needs to deliver capacity before the contracts become recognized revenue.
Nebius' customer commitments also depend on infrastructure delivery.
IREN's more than $4 billion year-end annualized run-rate revenue target is an operating target based on planned capacity, contracts, utilization and pricing rather than recognized accounting revenue today.
Applied Digital's long-term leases improve visibility but still require construction and financing.
Investors should therefore avoid comparing headline contract values directly with market capitalization without accounting for time, construction costs, financing expenses, depreciation and counterparty risk.
Power and Financing May Become the Next Bottlenecks
The next stage of the AI infrastructure cycle may be constrained as much by electricity and capital as by semiconductors.
Large grid connections can require years of development.
Companies that already control approved and connected power therefore have a strategic advantage.
Financing is the second constraint.
Hundreds of megawatts of AI infrastructure require billions of dollars, while computing hardware can become technologically obsolete relatively quickly.
If credit conditions tighten or investors begin demanding higher returns, project economics could deteriorate even if AI demand remains healthy.
The AI data center trade is therefore increasingly a technology, power, real estate and credit-market trade at the same time.
What the AI Data Center Rally Means for Crypto
IREN makes the connection between AI infrastructure and crypto particularly visible.
Bitcoin miners accumulated large portfolios of power, land, grid connections and data center expertise during the expansion of the mining industry.
Those assets are now being assigned an alternative value by AI customers.
If AI cloud contracts can produce higher and more predictable long-term returns from the same electricity, some miners may increasingly redirect capital from mining hardware toward AI computing infrastructure.
IREN is one of the clearest examples of that transition.
The effect on Bitcoin mining is not automatically positive.
Miners with high-quality power assets may gain a new valuation pathway, but AI operators also compete with miners for electricity, data center sites and investment capital.
Investors following both technology infrastructure and digital asset markets can use
MEXC to monitor changes in cross-asset risk appetite, although AI infrastructure equities and crypto assets continue to have fundamentally different cash flow and value-capture mechanisms.
Exclusive View from James Mitchell
The most important feature of this rally is not the size of CoreWeave's gain or the volatility in Nebius.
It is the widening of the AI valuation framework.
The first phase of the AI trade focused primarily on advanced GPU supply. The latest data suggest that accelerators are only one part of the constraint.
If CoreWeave's near-term capacity is effectively sold out, Nebius can sign increasingly valuable contracts, IREN can monetize former Bitcoin mining infrastructure through multi-year AI cloud agreements, and Applied Digital can secure long-duration leases for high-density data center capacity, scarcity has expanded across the infrastructure stack.
That is the structural signal behind the synchronized move.
The potential mistake is assuming that infrastructure scarcity means every AI data center equity should rise indefinitely.
The risk structure differs significantly across these businesses.
For CoreWeave, backlog conversion, capital returns and financing costs are critical. For Nebius, investors should monitor customer commitments, delivery execution and capital efficiency during rapid expansion. For IREN, the key question is whether new AI capacity achieves the pricing and utilization embedded in its run-rate targets. For Applied Digital, long-term returns depend on construction costs, financing and lease economics.
From a quantitative perspective, the correlation between these stocks is also worth monitoring.
If CRWV, NBIS, IREN and APLD remain highly correlated while backlog, utilization, cloud revenue and rental income all improve, the sector move has stronger fundamental support.
If share prices remain tightly correlated while operating performance starts to diverge, the rally may increasingly reflect thematic capital flows rather than business fundamentals.
Compute pricing is another important forward indicator.
Current contract economics suggest capacity remains scarce. At the same time, AI infrastructure providers are collectively spending enormous amounts to add new supply.
Today's shortage can become additional supply several years from now.
The long-term question is therefore not whether more AI data centers will be built. It is whether underlying demand can continue expanding faster than the capacity now being financed and constructed.
For crypto markets, IREN provides another useful signal.
Historically, the main operating question for a Bitcoin miner was how efficiently it could convert electricity into hash rate.
For some infrastructure owners, the decision may increasingly become whether one megawatt should support Bitcoin mining or AI computing.
That makes power infrastructure a common valuation variable connecting segments of the AI equity market with the crypto mining sector.
FAQ
Why are AI data center stocks surging?
AI data center stocks rallied after strong CoreWeave and Nebius results reinforced confidence in demand for computing infrastructure. CoreWeave reported approximately $104 billion of backlog, while Nebius AI cloud revenue increased 514% year over year. Investors responded by repricing not only cloud providers but also companies controlling data centers, power and deployment capacity.
Why did CoreWeave stock rise?
CoreWeave reported $2.575 billion of Q2 revenue, up about 112% year over year, and approximately $104 billion of revenue backlog excluding more than $25 billion of new commitments secured in early Q3. The company also raised its 2026 capital expenditure outlook to $35 billion to $39 billion as near-term compute capacity remained heavily committed.
Why did Nebius stock surge?
Nebius reported $582.3 million of Q2 revenue, up 454% year over year, while its core AI cloud business generated approximately $574.9 million, up 514%. The company also signed four major cloud contracts averaging more than $1 billion in total contract value and expects more than $9 billion in customer prepayments during 2026.
Why is IREN considered an AI data center stock?
IREN historically operated large-scale Bitcoin mining infrastructure but is increasingly redirecting its power and data center portfolio toward AI cloud services. The company has signed billions of dollars of AI contracts and raised its year-end 2026 AI cloud annualized run-rate revenue target to more than $4 billion, making AI infrastructure increasingly important to its valuation.
What is the relationship between Applied Digital and CoreWeave?
CoreWeave is a major tenant of Applied Digital's data center infrastructure. Applied Digital has disclosed 400 MW of contracted CoreWeave capacity at its Polaris Forge 1 campus, representing approximately $11 billion of prospective lease revenue. Strong CoreWeave demand can therefore improve confidence in the utilization and economic value of Applied Digital's infrastructure.
Are CRWV NBIS IREN and APLD the same type of company?
No. CoreWeave and Nebius are specialized AI cloud providers. IREN combines power infrastructure, data centers and AI cloud services and has a significant Bitcoin mining history. Applied Digital is more focused on developing and leasing physical data center capacity. They benefit from the same infrastructure cycle but have different revenue models and financial risks.
What are the biggest risks for AI data center stocks?
Major risks include excessive capital spending, higher financing costs, construction delays, insufficient power availability, customer concentration, rapid hardware depreciation and eventual compute oversupply. If newly constructed capacity grows faster than real AI demand, cloud pricing and data center returns could decline even if overall AI adoption continues to expand.
Can the AI data center boom affect Bitcoin miners?
Yes. AI data centers and Bitcoin mining compete for power, grid capacity, land and capital. Miners with high-quality infrastructure may be able to redirect some assets toward AI computing if long-term returns are more attractive. IREN already demonstrates this strategy, but greater AI demand can also increase competition for resources used by the Bitcoin mining industry.
Disclaimer
This content is provided for general informational and market research purposes only and does not constitute investment advice, financial advice, legal advice, tax advice or a recommendation to trade any security, cryptocurrency, derivative or other financial instrument. Cryptocurrencies, equities and related financial assets can experience substantial price volatility, and investors may lose part or all of their capital. Historical performance, technical indicators, financial metrics and on-chain data do not guarantee future results. Backlog, customer commitments, capital expenditure plans and annualized revenue targets may also change because of delivery schedules, financing conditions, customer demand and broader market conditions. Readers should conduct their own research and make decisions based on their financial circumstances, investment objectives and risk tolerance. The MEXC Crypto Pulse team accepts no responsibility for direct or indirect losses arising from the use of information contained in this content.
About the Author
James Mitchell specializes in technical analysis, market trends, and trading strategies for both Bitcoin and altcoins. Based in London, he has over 10 years of experience in financial markets. Before joining MEXC Learn, James worked as a senior analyst at a leading European investment firm, where he developed expertise in risk management and quantitative trading. His transition to cryptocurrency markets began in 2017, and he has since become recognized for his data-driven approach. He holds a Master's degree in Financial Economics from the London School of Economics. His analytical approach combines traditional technical analysis with on-chain metrics to provide readers with actionable insights.
Areas of Expertise:
Research References