Microsoft Azure

Neocloud provider CoreWeave reported revenue of $2.58 billion in the second quarter, more than doubling and topping analyst estimates due to high demand for AI infrastructure. The company's revenue backlog swelled 246% year-over-year to $104.2 billion, and the stock surged following the earnings call. CEO Michael Intrator stated that CoreWeave’s near-term capacity remains effectively sold out.

The article argues that U.S. dominance in AI stems from control of infrastructure like data centers, cloud computing, and AI servers, not just model performance. It highlights Nvidia's 85% market share in AI chips and its ecosystem, including partnerships with cloud providers and defense contracts, as central to American technological hegemony.

Enterprise cloud infrastructure spending reached $143 billion in Q2 2026, growing 43% year-on-year driven by AI adoption and demand for scalable IT solutions. Public IaaS and PaaS platforms grew 47%, with AI-specific cloud services surging 165%. AWS, Azure, and Google Cloud dominate 67% of the market, while tier-two providers like CoreWeave and Oracle show high growth.

Databricks is extending its partnership with Microsoft Azure into the 2030s, committing to run its core operations on Azure and increasing its use of Microsoft's Arm-based Cobalt custom processors.

Forrester's analysis states that despite Europe's investments in semiconductor manufacturing, it will remain dependent on US cloud providers and software. The report highlights the US and China as dominant in tech sovereignty, with European countries showing minimal progress in improving their tech sovereignty scores by 2030.

Vercel CEO Guillermo Rauch stated that companies are no longer relying on a single AI lab for all needs, instead using different labs for various parts of their AI stack. He highlighted the adoption of models from OpenAI, Anthropic, Gemini, DeepSeek, and GLM-5.2, emphasizing cost efficiency and performance.

Meta is exploring a cloud-computing business to sell unused AI computing capacity from its infrastructure, aiming to monetize surplus resources and offset AI investment costs. The potential service could compete with major cloud providers like Amazon Web Services, Microsoft Azure, and Google Cloud by offering raw compute or managed AI services.

Modern agentic AI infrastructure relies heavily on CPUs for managing data movement, workload isolation, and scheduling across distributed systems. Arm's expansion in datacenters, through collaborations like Meta's high-density CPU design and Arm Neoverse V3, highlights the growing importance of efficient, secure, and flexible CPU architectures. Major cloud providers (AWS, Google Cloud, Microsoft Azure) and NVIDIA are adopting custom Arm-based processors to optimize performance and efficiency in AI and cloud-native environments.

Google Cloud experienced network disruptions in India due to a fire at a third-party data centre, causing emergency shutdowns and elevated latency in Delhi, Chennai, and Mumbai. The outage impacted app and website performance, with no immediate workaround available as restoration efforts continued.

PCPJack has hijacked cloud servers from AWS, Google Cloud, and Azure to create a covert SMTP relay network. Compromised business servers in the U.S., Europe, and Asia were converted into SMTP proxies, verified for mail relay capability, and synced every five minutes.

Major cloud providers like AWS, Google Cloud, Microsoft Azure, and Oracle are adopting Arm-based processors to improve performance and energy efficiency in cloud infrastructure. Companies such as Spotify, Pinterest, and Uber report significant cost savings and performance gains using Arm-based instances, with industry trends shifting toward heterogeneous compute strategies.

A recent partnership between Microsoft and Anthropic could generate $43 billion in revenue for Microsoft Azure by 2030, according to HSBC. The news raises questions about Microsoft's potential shift in focus from OpenAI to Anthropic.

Big tech companies are making large profits but have unstable cash flows due to investing in the AI boom. Five major cloud companies are prioritizing the growth of AI over their own financial returns. This strategy may have long-term benefits but currently conceals unsettling financials.

ServiceNow has expanded its AI Control Tower to manage AI assets across an entire enterprise, including those outside its platform. The updated tower operates in five areas and includes 30 new enterprise connectors. This expansion aims to address AI agent sprawl and provide a command center for managing AI assets.

ConsentFix v3 is a new automated attack method targeting Microsoft Azure through OAuth abuse, leveraging hacker forums for distribution and enhancing previous techniques with automation and scaling capabilities.

Lawmakers and industry experts are debating whether data centers should be designated as a standalone critical infrastructure sector to enhance federal protection against cyber and physical threats. Recent attacks on Amazon data centers and growing reliance on cloud providers like AWS, Microsoft Azure, and Google Cloud have intensified the discussion, with the UK already recognizing data centers as critical infrastructure.

OpenAI and Microsoft have amended their partnership agreement, ending OpenAI's exclusive reliance on Microsoft's Azure cloud. Microsoft retains a non-exclusive license for OpenAI's IP until 2032, while OpenAI can now offer its models through other cloud providers. Revenue sharing terms have been adjusted, with payments capped and no longer tied to achieving artificial general intelligence (AGI).

The article highlights that hyperscalers (large tech companies) have already outspent most famous U.S. megaprojects, emphasizing their significant investment compared to traditional infrastructure projects. The piece is shared on Twitter and discussed on Hacker News with minimal engagement.

UK users are reporting capacity issues with Microsoft Azure, which is causing significant disruptions. The article notes that Sweden is being suggested as an alternative location for workloads due to these problems.