From Mining to Artificial Intelligence: A $9.1 Billion Contract Shifts Riot Platforms' Business Model

cryptonews.ruОпубліковано о 2026-08-15Востаннє оновлено о 2026-08-15

Анотація

The American bitcoin mining company Riot Platforms has entered into a major 20-year lease and service agreement for a 191 MW data center in Texas. The contract, reportedly with AI lab Anthropic, is expected to generate approximately $9.1 billion in revenue through 2048, potentially reaching $16.1 billion with extensions. This deal was announced alongside Riot's Q2 2026 results, which revealed a net loss of $237.2 million despite an 11% increase in bitcoin production to 1,587 BTC. While mining revenue fell 19.3% year-over-year to $113.7 million, overall revenue grew 14% to $174.2 million, boosted by a $23.2 million contribution from the data center segment. The contract signifies a strategic shift for Riot, providing a long-term revenue stream decoupled from bitcoin's price volatility. This move is part of a broader industry trend where bitcoin miners are repurposing energy infrastructure to serve the high-demand AI computing sector.

American mining company Riot Platforms has entered into a 20-year data center lease and service agreement, securing 191 MW of critical IT capacity at the Rockdale campus in Texas, with one of the leading AI labs. The agreement is effective until June 2048 and is expected to generate revenue of approximately $9.1 billion over the initial term. With two five-year extension options, the total contract value could reach roughly $16.1 billion.

According to Bloomberg citing sources familiar with the matter, the counterparty in the deal is Anthropic, the developer of the Claude neural network. Riot Platforms did not name the lessee in its official release, and Anthropic has not publicly commented on the deal.

Second Quarter Results: Bitcoin Sales and Loss

The agreement was revealed against the backdrop of Riot Platforms' financial results for the second quarter of 2026, ended June 30. During this period, the company sold about 4,300 bitcoins — the proceeds were used to fund operational activities and capital expenditures for data center expansion. The quarter ended with a GAAP net loss of $237.2 million, compared to a positive net profit of $219.5 million a year earlier.

Riot Platforms mined 1,587 bitcoins during the quarter — an 11% increase from 1,426 $BTC a year earlier. However, the average cost to mine one coin excluding depreciation rose to $49,912 compared to $48,992 in the second quarter of 2025. Despite the increase in mining volume, revenue from Bitcoin mining decreased by 19.3% year-over-year — to $113.7 million from $140.9 million. The company's total revenue, however, grew by 14%, to $174.2 million, due to a $23.2 million contribution from the data centers segment.

At the end of the quarter, Riot Platforms had 11,380 $BTC on its balance sheet, of which 5,821 $BTC were pledged as loan collateral. Based on the Bitcoin price of $58,527 as of June 30, 2026, the value of these assets was estimated at approximately $666 million. The company's liquid assets exceeded $1.2 billion, including $548.9 million in cash, of which $77.5 million were restricted funds.

The second quarter showed that Riot Platforms' classic mining business is facing revenue pressure even as production increases, while a net loss replaced last year's profit. Meanwhile, the contract with Anthropic opens a multi-year revenue stream for the company independent of Bitcoin price and mining costs.

AI Opinion

From the perspective of data-driven analysis, Riot Platforms' deal fits into a broader industry trend: Bitcoin miners are massively converting power capacity into assets for the AI industry. According to Bernstein's analysis, over the past two years, miners have contracted about 7 gigawatts of capacity with hyperscalers and AI cloud providers in deals worth over $135 billion across 19 transactions — the contract with Anthropic is far from the only one: in early July, TeraWulf signed a similar 20-year agreement for 401 MW at a campus in Kentucky.

The economics of this transition are not free: building AI data centers to meet hyperscaler requirements requires significant borrowed capital, while the margin profile of classic mining remains under pressure from rising production costs. What will happen to the balance sheets of mining companies if the demand for AI computing slows down before long-term infrastructure commitments are paid off?

Трендові криптовалюти

Пов'язані питання

QWhat is the core news announced by Riot Platforms, and with which company is the deal reportedly made?

ARiot Platforms announced a 20-year lease and maintenance contract for a 191 MW data center at its Rockdale, Texas campus. The deal is expected to generate approximately $9.1 billion in revenue over the initial term. According to sources cited by Bloomberg, the counterparty in the deal is Anthropic, the developer of the Claude AI model, though Riot did not officially name the tenant.

QWhat were Riot Platforms' key financial results for Q2 2026, particularly regarding Bitcoin mining and profitability?

AFor Q2 2026, Riot Platforms reported a net GAAP loss of $237.2 million, compared to a net profit of $219.5 million in Q2 2025. While Bitcoin production increased by 11% to 1,587 BTC, mining revenue fell 19.3% year-over-year to $113.7 million. The company also sold approximately 4,300 BTC to fund operations and expansion.

QHow does the new contract with Anthropic change Riot Platforms' business model and revenue streams?

AThe long-term contract with Anthropic provides Riot Platforms with a multi-year revenue stream that is not directly tied to Bitcoin price volatility or mining costs. This marks a strategic shift, diversifying the company's income away from reliance on classic Bitcoin mining, which is currently facing profitability pressures despite increased production.

QWhat broader industry trend does the Riot-Anthropic deal represent, and can you name another similar recent agreement?

AThe deal is part of a broader trend where Bitcoin miners are repurposing their energy infrastructure to serve the AI industry. Analysis by Bernstein indicates miners have contracted around 7 GW of power capacity to hyperscalers and AI cloud providers in recent years. Another example is the 20-year, 401 MW agreement signed by TeraWulf for a campus in Kentucky in early July.

QWhat are the potential risks associated with the industry's shift from Bitcoin mining to AI data center operations?

AThe transition involves significant costs, as building AI data centers requires substantial debt financing. A key risk is that if demand for AI computing slows down before the long-term infrastructure investments are paid off, it could strain the balance sheets of mining companies. The profitability of classic mining also remains under pressure from rising production costs.

Пов'язані матеріали

Pearl: Can You Simultaneously Conduct AI Inference and Mine Cryptocurrency? | Project Introduction

Pearl: Can AI Inference and Mining Really Be Done Simultaneously? A Project Introduction Pearl is a new proof-of-work (PoW) blockchain that aims to merge cryptocurrency mining with useful AI computation. Its core idea is to replace the traditional, arbitrary cryptographic hashing in PoW (like Bitcoin's) with matrix multiplications—the fundamental operation in neural network inference. In theory, a GPU performing these multiplications for an AI task (like generating text) could simultaneously be participating in block validation and earning the native PRL token. This proposes a "2-for-1" scenario where one unit of energy produces both AI output and cryptocurrency. The concept tackles the perceived "waste" in traditional PoW by making the computation useful. However, the critical challenge lies in ensuring this useful work is performed in a trustless, permissionless manner like Bitcoin. Pearl's protocol uses cryptographic techniques (adding verifiable "noise" to matrices and requiring proofs for small slices of the calculation) to prevent miners from cheating by using easily computable random numbers instead of genuine AI model weights and data. Despite the technical design, current network data suggests a significant gap between theory and practice. Analysis indicates the vast majority of the network's substantial computing power (tens of Exahashes) is still spent on multiplying randomly generated matrices, not on processing real AI inference requests. The "useful" output of the network is currently near zero. This happens because the protocol verifies the computational work (the multiplication) but cannot inherently verify the semantic usefulness or origin of the input data. Miners economically optimize for the block reward, not for producing usable AI results. Pearl's development includes releasing certified variants of open-source AI models (like Llama, Qwen) optimized for its system and partnerships (e.g., with Together AI) to offer discounted inference services subsidized by the future value of PRL tokens. The project bets that a future surge in scalable, low-latency AI inference demand will make it more profitable for miners to run real models than random data, aligning the network's security with productive output. Until then, it operates primarily as a GPU mining chain with a different, AI-compatible algorithm.

marsbit42 хв тому

Pearl: Can You Simultaneously Conduct AI Inference and Mine Cryptocurrency? | Project Introduction

marsbit42 хв тому

September US Bond Sell-off: Not Inflation Expectations, But Soaring Real Interest Rates

In September, the U.S. Treasury market experienced a significant sell-off. The yield on the 10-year Treasury surged from 4.75% to 5.29%, a 54 basis point increase. However, analysis reveals that over 90% of this rise, about 49 bps, came from a jump in real yields (as measured by TIPS), while market-implied inflation compensation rose only slightly by about 5 bps. This indicates the sell-off was driven primarily by a spike in real interest rates, not renewed inflation fears. This raises a key question: why is the market demanding higher real returns? Explanations could include stronger growth expectations, monetary policy adjustments, or higher risk premiums for holding long-term bonds. Notably, this bond market stress coincided with strong performance in AI-related tech stocks, particularly semiconductors, while other rate-sensitive sectors declined. This divergence suggests investors are pricing in vastly different futures for different assets. QuantStreet Capital's Harry Mamaysky proposes one explanation: the market may be repricing for both stronger future economic growth and the massive capital demands from Big Tech's AI infrastructure expansion. In this view, higher real rates and strong AI stocks aren't contradictory if investors believe future AI-driven profit growth can offset higher financing and discount rates. Yet, this remains a theory. Rising real yields could also reflect an increase in the term premium—the extra compensation investors require for holding long-term bonds—due to fiscal concerns or market volatility, rather than pure growth optimism. This distinction is crucial: if driven by growth, higher rates can coexist with rising equity valuations; if driven by a higher term premium, it could pressure corporate profits and asset prices more broadly. The critical unresolved issue is whether the market's demand for higher real returns stems from confidence in future economic profits (especially from AI diffusion) or from a reassessment of the risks of holding long-term assets. The answer will define the path for bonds, stocks, and the ongoing AI investment cycle.

marsbit51 хв тому

September US Bond Sell-off: Not Inflation Expectations, But Soaring Real Interest Rates

marsbit51 хв тому

Goldman Sachs Research Report Interpretation: Dell Delivers 4 to 6 Weeks Faster Than Peers, Memory Is the 2027 Bottleneck

Goldman Sachs Research Report Summary: Dell's Competitive Edge and Future Outlook (Based on October 8, 2026 Report) Dell maintains a competitive advantage in AI server delivery, with a lead time approximately 4-6 weeks faster than its competitors. This speed-to-market allows customers to generate revenue from their infrastructure sooner, supporting Dell's premium pricing relative to other OEMs and ODMs. Goldman Sachs reiterates a Buy rating with a $570 price target. Key points from the report, based on a management meeting at Dell's headquarters, include: * **Extended Competitive Moat:** Dell's leadership extends beyond component procurement to differentiated capabilities in system design, supply chain, manufacturing, and deployment. The company's scale and pre-integration testing enable faster deployment of large clusters. * **2027 Memory Bottleneck:** Supply remains a primary constraint for AI infrastructure growth. Dell expects memory supply tightness to persist into 2027, with demand growth outpacing supply improvements. Its scale and direct sales model provide early visibility into demand shifts, aiding in prioritizing allocations for AI/server opportunities over weaker segments like PCs. * **Profitability Drivers:** Future margin expansion is expected to be driven primarily by storage solutions (leveraging Dell's own IP like PowerScale) and operational leverage. AI server margins are maintained in the mid-single-digit range, seen as a balance between profitability and competitive positioning. Component cost fluctuations, especially for memory, are passed through to customers. * **Enterprise AI Adoption:** Enterprise AI adoption is still in early stages. Dell believes successful deployment requires data preparation and business process redesign, not just compute infrastructure. A hybrid architecture model is anticipated, combining cloud-based cutting-edge models with on-premise open-source models for cost control and security. The report concludes that the sustainability of Dell's premium pricing hinges on maintaining its delivery speed, order quality, and storage attach rates, with 2027 memory supply being a critical near-term factor to watch. *Disclaimer: This summary is based on a third-party Goldman Sachs research report dated October 8, 2026, and is for informational purposes only.*

marsbit1 год тому

Goldman Sachs Research Report Interpretation: Dell Delivers 4 to 6 Weeks Faster Than Peers, Memory Is the 2027 Bottleneck

marsbit1 год тому

U.S. Treasury Yields Soar, Why Can AI Stocks Continue to Rise?

In September, U.S. financial markets experienced a notable divergence: while long-term Treasury yields surged dramatically—the 10-year yield rising over 50 basis points—and fixed-income assets declined, AI-related and semiconductor stocks continued to rally, with the Nasdaq holding relatively firm. In contrast, small and mid-cap stocks, the equal-weight S&P 500, and rate-sensitive sectors lagged. This dynamic raises a key question: if higher interest rates typically pressure equity valuations by raising discount rates, why have tech stocks remained strong? The article, referencing analysis from QuantStreet Capital, suggests one interpretation: markets may be pricing in sufficiently high future earnings growth from AI, potentially enough to offset the valuation drag from elevated rates. However, a deeper puzzle emerges. The market optimism appears concentrated in AI infrastructure suppliers like semiconductor firms, which directly benefit from soaring capital expenditure (Capex) on data centers and chips. There is little evidence yet of this enthusiasm spreading to the broader corporate sector (the "Rest of the Corporate Sector"), whose future profits are supposed to ultimately justify these massive AI investments. This highlights the critical uncertainty in the AI investment thesis: who ultimately captures the economic value? While chipmakers see near-term revenue, the long-term sustainability depends on whether AI adoption boosts productivity and profitability for the wider economy, enabling end-user companies to bear the high costs. Preliminary productivity data shows some improvement, but direct evidence of AI-driven profit gains outside the tech sector remains sparse. Moving forward, key signals to watch include: the sustainability of AI infrastructure spending, evidence of AI translating into improved profitability for non-tech companies, and whether the rise in long-term yields reflects strong growth expectations or other risks like fiscal concerns. The current market behavior suggests confidence in the AI builders, but the broader payoff from AI—essential for justifying current valuations—is still pending confirmation.

marsbit1 год тому

U.S. Treasury Yields Soar, Why Can AI Stocks Continue to Rise?

marsbit1 год тому

Торгівля

Спот

Популярні статті

Як купити DATA

Ласкаво просимо до HTX.com! Ми зробили покупку DATA Network (DATA) простою та зручною. Дотримуйтесь нашої покрокової інструкції, щоб розпочати свою криптовалютну подорож.Крок 1: Створіть обліковий запис на HTXВикористовуйте свою електронну пошту або номер телефону, щоб зареєструвати обліковий запис на HTX безплатно. Пройдіть безпроблемну реєстрацію й отримайте доступ до всіх функцій.ЗареєструватисьКрок 2: Перейдіть до розділу Купити крипту і виберіть спосіб оплатиКредитна/дебетова картка: використовуйте вашу картку Visa або Mastercard, щоб миттєво купити DATA Network (DATA).Баланс: використовуйте кошти з балансу вашого рахунку HTX для безперешкодної торгівлі.Треті особи: ми додали популярні способи оплати, такі як Google Pay та Apple Pay, щоб підвищити зручність.P2P: Торгуйте безпосередньо з іншими користувачами на HTX.Позабіржова торгівля (OTC): ми пропонуємо індивідуальні послуги та конкурентні обмінні курси для трейдерів.Крок 3: Зберігайте свої DATA Network (DATA)Після придбання DATA Network (DATA) збережіть його у своєму обліковому записі на HTX. Крім того, ви можете відправити його в інше місце за допомогою блокчейн-переказу або використовувати його для торгівлі іншими криптовалютами.Крок 4: Торгівля DATA Network (DATA)Легко торгуйте DATA Network (DATA) на спотовому ринку HTX. Просто увійдіть до свого облікового запису, виберіть торгову пару, укладайте угоди та спостерігайте за ними в режимі реального часу. Ми пропонуємо зручний досвід як для початківців, так і для досвідчених трейдерів.

763 переглядів усьогоОпубліковано 2026.07.01Оновлено 2026.07.01

Як купити DATA

Обговорення

Ласкаво просимо до спільноти HTX. Тут ви можете бути в курсі останніх подій розвитку платформи та отримати доступ до професійної ринкової інформації. Нижче представлені думки користувачів щодо ціни DATA (DATA).

活动图片