Core Scientific has reported a sharp rebound in its second-quarter financial performance as its data-center colocation business—built to support artificial intelligence (AI) and high-performance computing (HPC)—continues to drive results after the miner’s shift away from a Bitcoin-only model.

In earnings released Tuesday, the company said Q2 revenue rose to $164.2 million, compared with $78.6 million in the same quarter a year earlier. Colocation revenue made up the overwhelming majority of that figure, climbing to $136.7 million from $10.6 million, while gross profit increased to $70 million from $5 million.

Key takeaways

• Core Scientific’s revenue more than doubled in Q2, with colocation now the dominant earnings engine.
• AI- and HPC-oriented infrastructure appears increasingly central to the company’s profit trajectory, as gross profit jumped alongside colocation revenue.
• Despite strong topline growth, Core Scientific posted a large net loss driven largely by a non-cash accounting impact tied to warrant valuation.
• The company’s newly announced AMD partnership could support up to 2.5 GW of leasable capacity, with initial multi-site agreements beginning in 2027.

Revenue surge driven by colocation, not mining

The company’s results highlight how quickly Core Scientific’s operating profile has changed. According to the earnings figures, colocation revenue—rather than mining-related activity—accounted for $136.7 million of the quarter’s total $164.2 million. In the year-ago period, colocation contributed only $10.6 million, underscoring the scale of the pivot and the speed at which the business ramped.

Gross profit also rose substantially, reaching $70 million from just $5 million. While revenue growth alone can sometimes reflect mix effects or transitional capacity, the gross profit jump suggests Core Scientific’s shift is beginning to translate into a more favorable economics profile for its core infrastructure operations.

Core Scientific is no longer positioning itself as a pure-play Bitcoin miner. Earlier coverage from Cointelegraph noted that it generates the bulk of its revenue from colocation services while holding a comparatively small Bitcoin treasury of fewer than 1,000 BTC, based on industry data compiled by bitcointreasuries.net.

The net loss: accounting effects, not necessarily cash stress

Even as revenue and gross profit climbed, Core Scientific still recorded a $1.15 billion net loss. The company attributed the result primarily to a non-cash accounting charge connected to the rising value of outstanding warrants as its share price increased.

This matters for readers because the market often interprets net losses as immediate operational distress. Here, the earnings disclosure frames the loss as largely accounting-driven rather than a direct signal that the business is consuming cash faster than it generates it. In the context of a company transitioning to longer-term infrastructure contracts, that distinction can influence how investors evaluate near-term headlines versus underlying demand and contracted capacity.

Following the earnings release, Core Scientific’s shares reportedly fell by more than 4%, trimming its year-to-date gains—an indication that some investors may have focused on the net loss headline before digging into what drove it.

An AMD deal aims to lock in large-scale AI compute capacity

Alongside its quarterly results, Core Scientific announced a partnership with Advanced Micro Devices (AMD). AMD designs CPUs and AI-oriented graphics processors that compete with other major chip vendors.

The agreement is structured to support up to 2.5 gigawatts of leasable data-center capacity. The initial phase is anchored by 15-year agreements covering 530 megawatts across multiple US sites starting in 2027, with the ability to expand over time.

Core Scientific said the broader AMD partnership could generate more than $14 billion in contracted base revenue. The company also stated that its total leased customer power capacity is now roughly 1.1 GW, representing more than $24 billion in potential contracted revenue.

From an investor perspective, this type of power-and-capacity contracting is often viewed as a way to stabilize revenue in infrastructure businesses, especially when the demand side is tied to large compute requirements from AI training and inference workloads. For traders and equity holders, the key question becomes how quickly these longer-dated commitments translate into actual utilization and incremental margins—especially as the market moves from “plans” to “running load.”

Broader AI data-center competition signals shifting priorities across crypto infrastructure

Core Scientific’s quarter and its AMD partnership arrive as other infrastructure providers tied to the crypto era also expand into AI compute. Earlier this month, IREN disclosed $2.8 billion in cloud contracts with AI developers. Separately, Hut 8 unveiled a $9.8 billion lease agreement with an unnamed customer for capacity at its AI data campus.

Set against those moves, Core Scientific’s results look less like a standalone turnaround story and more like part of a sector-wide reallocation of resources. Bitcoin mining companies that secured data-center assets and power access during the mining buildout are increasingly competing on hosting, leasing, and compute-adjacent services rather than relying solely on block rewards.

Still, uncertainty remains. While contracted capacity figures and partnership announcements can support a longer-term growth narrative, the market continues to watch for execution details: how fast customers ramp usage, whether contracted power translates into sustained gross margins, and how balance-sheet dynamics—such as the accounting treatment of warrants—can affect headline profitability.

Investors should watch Core Scientific’s next reporting period for two things: whether the revenue mix continues to lean further into colocation and how management’s guidance and utilization metrics evolve as AMD-linked capacity approaches the initial 2027 ramp-up window.

Ten European financial institutions have formed a new jointly owned blockchain cooperative called Regulated Layer One (RL1), positioning it as a permissioned network for tokenized assets and regulated market infrastructure.

RL1 announced that it has been established as a European Cooperative Society in Luxembourg and has started operations with founding members including ABN AMRO, Cecabank, Chartered Investment, Crédit Mutuel Alliance Fédérale, DekaBank, DZ BANK, LBBW, Natixis CIB, SC Ventures and Seturion. The group says governance will be shared, with each member holding equal decision-making rights over the network’s development.

Key takeaways

• RL1 is launching as a Luxembourg-based European Cooperative Society with 10 founding financial institutions.
• The network is permissioned and aimed at institutional use cases such as tokenized bonds, collateral, and settlement.
• RL1 is built on infrastructure originally developed by German fintech Secure Worldwide Interbank Asset Transfer (SWIAT), now transferred to the cooperative.
• SWIAT reports more than 50 transactions worth over €700 million during three years of production use.
• Additional institutions are being discussed for membership, including NatWest, as RL1 begins operations.

A cooperative model for regulated blockchain infrastructure

RL1’s launch reflects a broader push among banks and other regulated players to build shared blockchain rails that can integrate with existing compliance and oversight frameworks. By organizing the network as a cooperative, RL1 is attempting to shift control away from single-operator models and toward governance shared across member institutions.

The founding structure matters for investors and market participants because governance can directly affect roadmap priorities—such as which tokenized asset standards are supported, how settlement workflows are designed, and how risk controls are maintained. RL1 says each member will have equal decision-making rights, signaling that the network is meant to evolve through collective agreement rather than unilateral changes.

From SWIAT-built infrastructure to RL1 ownership

RL1’s technical foundation traces back to infrastructure developed by German fintech Secure Worldwide Interbank Asset Transfer (SWIAT). According to the cooperative, SWIAT has transferred ownership of the network to RL1, marking a clear step from an originating build phase into an operator-and-governance phase under the cooperative structure.

SWIAT also provided performance context from its production use period, stating that the platform processed more than 50 transactions totaling over €700 million (about $808 million) across three years. While that figure is not presented as a measure of network scale in terms of daily volume or active users, it does frame the initiative as having moved beyond prototypes into live transaction processing before the cooperative launch.

Use cases designed for tokenization and settlement

RL1 says the blockchain is intended to support institutional workflows tied to digital money and tokenized financial instruments. The cooperative highlighted use cases including tokenized bonds, collateral management, and blockchain-based settlement.

One theme RL1 emphasizes is reducing fragmentation in distributed ledger efforts. Financial institutions have historically pursued separate DLT systems, often leading to interoperability challenges and duplication of integration work. RL1 argues that a shared network among regulated participants can mitigate those issues by giving members a common infrastructure layer for tokenized settlement-related activities.

For market participants, that framing matters because interoperability and shared settlement are frequently cited barriers to wider adoption of tokenized assets. A network that centralizes governance and standards among a defined group of regulated institutions can shorten the path to operational alignment—though it cannot eliminate the need for external integrations where assets or counterparties sit outside the network.

Leadership and expansion plans

RL1 named Henning Vollbehr, formerly Managing Director at SWIAT, as its leader. The cooperative also said that KfW and L-Bank will continue supporting the initiative, indicating the project retains institutional backing as it transitions into ongoing operations.

RL1 is also looking outward: the group said it is in discussions with additional institutions about joining the network, including NatWest. That expansion effort will likely be a key indicator of whether RL1 can grow beyond the initial consortium and increase its usefulness as a settlement and tokenization venue for more participants.

Going forward, readers should watch how RL1’s cooperative governance translates into concrete product decisions—especially around asset types, settlement rules, and interoperability with external systems—alongside whether the membership talks broaden participation beyond the founding banks.

Core Scientific shares fell after the Bitcoin miner agreed to provide AMD with up to 2.5 gigawatts of data center capacity for artificial intelligence deployments starting in 2027.

AMD secures up to 2.5 GW from Core Scientific

Core Scientific and AMD have signed an agreement covering up to 2.5 GW of data center capacity for customers deploying the chipmaker’s AI systems.

Capacity will become available from 2027, according to a joint announcement from the companies. Core Scientific and AMD will also work together on the physical design of the infrastructure needed to support high-density computing workloads.

Planned deployments will use AMD Instinct graphics processing units, EPYC processors, and the company’s ROCm software platform. The announcement did not disclose the financial value of the agreement or identify the end customers expected to use the capacity.

Unlike a standard hardware order, the arrangement pairs AMD’s computing products with Core Scientific’s power and data center infrastructure. The scale of the agreement could make AMD an important customer and commercial partner as Core Scientific converts more of its sites from crypto mining to AI computing.

AMD will also receive market-priced warrants allowing it to purchase Core Scientific common stock. The warrants remain subject to commercial conditions, and the companies did not disclose the potential size of AMD’s resulting stake.

Core Scientific accelerates its shift from Bitcoin mining

Core Scientific built its business around Bitcoin mining but has increasingly redirected capital and power capacity toward high-density data center services.

As previously reported by crypto.news, the company sold 2,385 Bitcoin earlier in 2026 to provide liquidity during the transition. BitcoinTreasuries data shows that Core Scientific still holds 848 BTC.

The company continues to generate revenue by mining crypto for its own account and providing hosting services to other miners. However, it is repurposing its remaining facilities for colocation services capable of supporting power-intensive AI systems.

The AMD agreement places Core Scientific among several publicly traded Bitcoin miners pursuing AI infrastructure contracts. Limited access to large sites with substantial power connections has made miners’ existing facilities attractive to cloud providers and AI developers.

MARA recently expanded its AI infrastructure footprint through the acquisition of a site in Texas. TeraWulf also signed a 20-year data center agreement with Anthropic earlier in July.

Hut 8 and IREN announced separate multibillion-dollar AI infrastructure deals last week. Hut 8 signed a second 15-year lease worth $9.8 billion at its Beacon Point campus in Texas, while IREN disclosed $2.8 billion in new multiyear AI cloud contracts.

CORZ reverses its premarket gain

CORZ initially rose more than 5% in premarket trading after the AMD agreement was announced. The stock reversed direction after the opening bell and fell more than 4% as a broader equity market sell-off weighed on trading.

Core Scientific shares have now declined more than 12% over the past week. Despite the latest pullback, the stock remains up over 40% since the start of 2026 as investors assess its transition from Bitcoin mining to AI infrastructure.

The reversal indicates that investors are weighing the agreement’s long-term capacity against near-term execution costs and market conditions. The companies did not disclose expected revenue, construction spending, deployment stages, or a timetable for bringing the full 2.5 GW online.

For U.S. investors, AMD’s warrants introduce a potential dilution consideration if the chipmaker exercises its right to buy CORZ shares. The commercial conditions and number of shares covered will determine the eventual effect on existing holders.

Execution becomes the next test for Core Scientific

Core Scientific must now prepare its facilities for deployments scheduled to begin in 2027. Its progress will depend on power availability, construction timelines, customer demand, and the capital required to convert former mining sites.

Investors will also watch for disclosures covering the agreement’s financial value, deployment schedule, and warrant terms. These details will help determine how quickly the AMD partnership could replace declining reliance on Bitcoin mining revenue.

The wider shift among miners is increasing competition for AI tenants and financing. Core Scientific’s 2.5 GW agreement gives it substantial contracted demand, but future results will depend on how much capacity is delivered and how profitably the company operates it.

Cloud mining platforms gain popularity in 2026 as investors seek simpler ways to access Bitcoin, DOGE, and LTC mining without expensive hardware.

Want to participate in Bitcoin mining in 2026 but don’t want to buy expensive mining rigs? Then cloud mining platforms remain one of the simplest and most worry-free options. 

Nowadays, more and more investors are entering the market through free cloud mining platforms, easily participating in mining mainstream cryptocurrencies such as BTC, DOGE, and LTC without needing to build their own equipment or bear high electricity and maintenance costs.

However, while there are many platforms on the market, only a few are truly worth considering. A good cloud mining platform should not only have a clear and transparent profit mechanism, but also a stable data center, an automatic payment system, and a sufficiently secure operational background.

Based on the market trends and platform characteristics in 2026, SHRMiner, BitFuFu, IQMining, Binance Cloud Mining, and CCG Mining are the five platforms that deserve close attention.

1. SHR Miner: The most noteworthy cloud mining platform in 2026

For those who are looking for a service that balances security, flexibility, and beginner-friendliness, SHRMiner is a very popular choice. Launched in 2018 and headquartered in the UK, SHRMiner operates over 100 large-scale renewable energy mining farms in the US, UK, Russia, Switzerland, Iceland, Virginia, Georgia, Vancouver, Canada, and other locations, utilizing renewable energy sources such as hydropower and wind power to enhance mining efficiency.

The platform supports mining mainstream cryptocurrencies such as BTC, LTC, and DOGE. Users do not need to purchase any hardware; they only need to select a suitable contract to start. Its contract coverage is extensive, with a comprehensive range of entry-level and premium packages to suit users with different budgets.

SHRMiner Core Advantages:

Register to receive a $15 bonus and free mining experience.

• Zero learning curve: No technical skills, hardware, or complicated operations required — just click to start mining.
• Supports daily automatic settlement, with no transaction fees or maintenance costs.
• Uses advanced ASIC mining equipment, connected to green energy, improving operational efficiency.
• Provides SSL encryption and DDoS protection.
• Provides a real-time earnings dashboard, allowing users to track their earnings anytime, anywhere.
• Supports multiple contract types including BTC, LTC, and DOGE.

SHRMiner gained popularity in 2026 primarily because it was suitable for beginners to quickly get started while also supporting more advanced users for flexible configuration. Its overall performance was well-balanced, from the initial user experience to contract scalability.

2. BitFuFu: A professional platform backed by Bitmain

BitFuFu has garnered significant market attention due to its association with Bitmain. This type of platform is particularly attractive to users who value mining rig resources and hardware expertise. BitFuFu is suitable for investors seeking a more mature mining service system.

3. IQMining: A key focus for long-term contract users

IQMining has been operating for several years and is characterized by offering longer-term mining contracts. For users who prioritize long-term planning over short-term volatility, IQMining is a common choice.

4. Binance Cloud Mining: Integrated trading and mining

The biggest advantage of Binance Cloud Mining lies in its ecosystem integration. Users can manage mining and asset transfers directly within their Binance accounts, eliminating the need for frequent platform switching. This is especially convenient for existing Binance investors.

5. CCG Mining: A key platform in the European market

CCG Mining offers a comprehensive range of services, including cloud mining, mining rig sales, and hosting. It enjoys considerable brand recognition in the European market and is suitable for users looking to explore diverse mining services.

Why are more and more people choosing cloud mining in 2026?

Compared to traditional mining rigs, the biggest advantages of cloud mining are:

• No need to purchase expensive equipment
• No need to bear high electricity bills
• No need for technical maintenance knowledge
• Quick start after registration

Some platforms also offer free trials and reward mechanisms. For ordinary users, this model is obviously more convenient and more suitable for low-barrier entry into the crypto market.

Conclusion: Which cloud mining platform is worth paying attention to in 2026?

From an overall user experience perspective, SHRMiner remains one of the most competitive platforms in 2026. It excels in platform transparency, mining process, settlement efficiency, and beginner-friendly features, while supporting multiple cryptocurrencies including BTC, LTC, and DOGE, making it highly versatile.

Of course, for those who prioritize exchange integration, Binance Cloud Mining will be more convenient; for those who value long-term stable contracts, IQMining and CCG Mining are also good options.

In general, when choosing the best cloud mining platform, it is recommended to focus on the platform’s background, security mechanisms, contract flexibility, and actual user experience. For users looking to start their free cloud mining journey in 2026, prioritizing a transparent, secure platform with clear settlement is a safer bet.

In short, for those who are looking for a cloud mining platform that balances transparency, flexibility, and ease of use in 2026, SHRMiner is a wise choice.

For more platform information, service details, and cloud computing solutions, please visit the official platform or download the mobile application.

Crypto theft and fraud losses exceeded $1 billion in the first half of 2026, according to Blockaid’s H1 2026 security report released Tuesday. The period also featured the highest number of hacks in any six-month stretch recorded by the onchain security firm.

Ethereum and Solana accounted for the largest portions of stolen funds, with approximately $332 million and $326 million respectively. Blockaid tracked 212 security incidents during the six-month period, including a standout single exploit tied to KelpDAO, which Blockaid reported at $292 million.

Key takeaways

• Blockaid estimates total crypto losses above $1 billion in H1 2026, alongside the highest six-month hack count in its historical data.
• Ethereum led by stolen-fund impact (about $332 million), largely driven by code and application-layer exploits.
• Solana’s losses were similarly high (about $326 million) but were overwhelmingly linked to compromised keys and signing infrastructure.
• The largest single incident in the report involved KelpDAO, with losses of $292 million.
• Blockaid reports high-threshold exploit verification increased sharply in H1 2026 compared with all of 2025.

Ethereum’s losses underline application-layer risk

Blockaid said Ethereum incurred the highest losses from incidents in H1 2026, with attackers primarily focusing on vulnerabilities in applications built on the network. By count, code exploits were the dominant driver of Ethereum incidents, and Blockaid also highlighted that several major loss events involved compromised keys.

Among the notable incidents cited in the report were the Humanity Protocol and StablR attacks. CoWSwap was singled out as the only major Ethereum incident categorized as a user mistake, rather than a protocol or code vulnerability.

Blockaid described recurring techniques behind Ethereum-related breaches, including flaws in bridges and smart contracts, unauthorized access to privileged accounts, and market manipulation methods. While these categories differ in mechanics, they share a common theme: high-value Ethereum apps present a dense target surface for attackers seeking direct exploitation paths and privileged access.

The report also emphasized Ethereum’s role as a hub for major crypto primitives—restaking platforms, stablecoins, and decentralized exchanges—where substantial capital and complex integrations can concentrate both the value at risk and the probability of exploitable edge cases.

Solana’s stolen funds spiked with a shift toward key compromises

Solana’s losses in the first half of 2026 nearly matched Ethereum’s. Blockaid estimated stolen funds around $326 million for the period, a clear jump from roughly $127 million recorded in 2025.

In an observation carried in the report, Blockaid CEO Ido Ben-Natan told Cointelegraph that 2025 totaled $2.58 billion lost across 63 incidents, with activity concentrated in Q1 and with Ethereum and Arbitrum topping stolen-fund flow at the time.

However, Blockaid said Solana’s deterioration in H1 2026 did not come from a surge in smart contract exploits. Instead, compromised keys accounted for more than 98% of Solana’s losses, which Blockaid linked largely to incidents involving Drift Protocol and Step Finance. Blockaid also attributed those events to North Korea-linked cyber groups.

This matters because it reframes the operational priorities for Solana-related infrastructure. While Ethereum incidents in the report were more closely tied to vulnerabilities in protocol code, Blockaid said Solana-related losses were primarily associated with signer infrastructure and organizational security failures. In other words, the dominant threat vector in this period was not “bugs in execution,” but failures in control systems and signing operations.

Blockaid noted that only a small fraction of Solana losses were tied to code exploits—citing Raydium and Volo as examples—underscoring how heavily the report’s Solana narrative depends on key and signing security rather than on-chain contract defects alone.

What’s changed in the threat landscape

Two shifts stand out across Blockaid’s findings for H1 2026. First, Ethereum’s risk profile remained oriented around smart contract and application-layer weaknesses, where bridges, contract logic, and privileged account access can be exploited. Second, Solana’s losses, despite being comparable in size to Ethereum’s, were driven overwhelmingly by compromised keys and signing infrastructure—an operational and security governance problem rather than a pure software vulnerability issue.

Blockaid also reported that it verified 3.4 times as many high-threshold exploits in H1 2026 as it did across all of 2025. That suggests either that attackers pursued more severe, high-confidence exploitation paths during the period, or that the environment—across targets and integrations—supported higher-impact outcomes. In practice, for teams defending networks and protocols, it raises the likelihood of facing fewer “small” issues and more attacks with direct paths to material loss.

Finally, the report’s largest single exploit—KelpDAO at $292 million—fits the broader pattern of high-value targets attracting concentrated attacks. Even when overall incident counts vary, a small number of high-impact events can dominate the stolen-fund totals, which is precisely what appears in Blockaid’s H1 2026 breakdown.

Closing perspective

As H1 2026’s losses show, the most consequential threats are not uniform across chains: Ethereum-focused defenses should emphasize application and privileged access security, while Solana stakeholders should treat key management and signing infrastructure resilience as top priority. Readers should watch whether the disparity between code-driven incidents and key-driven incidents persists in the second half of 2026, and whether incident counts remain elevated alongside high-threshold exploit activity.