Cloud · Top stories
Google Expands Gemini Enterprise Agent Platform with Remote MCP Server
Google has enhanced its Gemini Enterprise Agent Platform by introducing a remote Managed Control Plane (MCP) server. This update allows developers to securely connect external AI agents with Google Cloud resources, facilitating agent development across various IDEs. The enhancements address developer feedback on building more efficient, production-ready AI agents.
SpaceX Aims for Orbital Data Centers with Plan for 1 Million Satellites
SpaceX, led by Elon Musk, has filed for an FCC application to create an orbital data center constellation of up to 1 million satellites. The project faces technical challenges such as satellite launch, heat dissipation, and radiation management. If successful, these centers could power millions of GPUs, potentially transforming data center operations. However, achieving the scale and timeline remains uncertain due to current launch and manufacturing capacities.
Google Acquires Crusoe's Wyoming Data Center Site, Plans $17 Billion Investment
Google has acquired a data center site in southeast Wyoming from Crusoe, which had completed pre-development work including site selection, governmental approvals, and a hybrid power strategy. Google plans to invest $17 billion to develop the campus, taking over the project after reportedly raising concerns about its cost and timeline under Crusoe's management. Crusoe will continue to develop another data center for Google near Amarillo, Texas.
Iranian missiles reportedly destroy Amazon's Bahrain data center amid conflict escalation
Iranian state media claims that the IRGC struck and destroyed Amazon's Bahrain data center with cruise missiles. This attack is stated to be retaliation against U.S. strikes on an Iranian nuclear facility, highlighting ongoing threats to American tech infrastructure in the region.
Meta to Launch Cloud Business, Enter AI Compute Market
Meta is preparing to launch a cloud infrastructure business to sell excess AI computing power and models. This strategy would place Meta in direct competition with AWS, Google Cloud, and Microsoft Azure and aims to monetize its significant AI investments, improving profitability. Reports of early talks with Anthropic for leasing compute power highlight potential partnerships in the cloud service sector.
ReadyOn Details Four-Layer Tenant Isolation Architecture for Multi-Tenant Amazon EKS
ReadyOn implemented a four-layer tenant isolation architecture for its Harmony platform on Amazon EKS, combining Kubernetes namespaces, dedicated Karpenter node pools, Amazon VPC security groups, and per-tenant Amazon Aurora databases. This design aims to prevent cross-tenant access at the Kubernetes, compute, network, and data layers, addressing security challenges in multi-tenant environments.
Google Open-Sources TorchTPU for Native PyTorch on TPUs, Hugging Face Integrates
Google has open-sourced TorchTPU, a native PyTorch backend, enabling PyTorch users to run models directly on Google's Tensor Processing Units (TPUs) with minimal code changes. This integration makes TPUs a native device option alongside GPUs and CPUs for PyTorch workloads, simplifying development and deployment. Hugging Face has also integrated TorchTPU, allowing models from the Hugging Face Hub to run natively on TPUs.
Southeast Asian Datacenter Boom Slows Amid Regulatory and Environmental Concerns
Thailand paused 49 datacenter construction projects and delayed decisions on 117 more due to insufficient regulatory details, while Indonesia halted a project over permits and environmental assessments. This slowdown occurs as Southeast Asia leads the Asia Pacific region in under-construction datacenter capacity, driven by global AI demand.
Texas Halts New Data Center Grid Connections for Audit Amid Power Demand Concerns
Texas Governor Greg Abbott has ordered a moratorium on new data center approvals and grid connections through the Public Utility Commission of Texas (PUCT) and the Electric Reliability Council of Texas (ERCOT) until an audit is completed. This action affects over 1,800 projects, primarily data centers, representing 474 gigawatts of potential demand, which is five times the state's record peak demand and raises concerns about grid stability.
Amazon warns against blocking data centers, citing risks to US economy and national security
Amazon Web Services CEO Matt Garman published a blog post urging communities to support AI data center projects, arguing that blocking them could harm the US economy and national security. The post addresses public concerns regarding data centers' impact on jobs, power, and the environment, while also announcing a "Data Center Commitment" to invest in communities and end nondisclosure agreements with government agencies.
Finnish Authorities Halt Google Data Center Preparatory Work Over Environmental Concerns
Finnish authorities have ordered Google to suspend preparatory work at two planned data center sites in Muhos and Kajaani, Finland, due to the absence of mandatory environmental impact assessments (EIAs). The Finnish Supervisory Agency (LVV) issued the order after Google's subsidiary, Tuike Finland Oy, cleared over 300 hectares of forest and performed other site alterations without the required assessments. Google acknowledged it "fallen short of our own high standards" and will comply with the agency's guidance, impacting its €13bn investment in Finland.
Amazon Seeks FCC Approval for 5,105-Satellite Direct-to-Device Network by 2028
Amazon has filed an application with the Federal Communications Commission (FCC) to launch a new constellation of 5,105 low Earth orbit (LEO) satellites, with deployment planned to begin in 2028. This initiative aims to provide direct-to-device services for voice, messaging, data, and emergency services, leveraging Globalstar's spectrum and integrating with Amazon's existing satellite infrastructure. The network will target users in areas unserved or underserved by traditional wireless providers.
Microsoft to Deploy AMD's Helios AI Accelerator on Azure
Microsoft will integrate AMD's Helios, a rack-scale AI accelerator, into its Azure cloud infrastructure. This partnership aims to enhance AI computational power for Azure users, directly competing with Nvidia's offerings. The deployment underscores the growing demand for AI compute capabilities.
AWS Billing Bug Leads to Erroneous Billion-Dollar Estimates for Customers
AWS users experienced incorrect billing estimates due to a bug causing charges to show as billions of dollars instead of reflecting actual usage. The fault in the AWS billing computation system involved incorrect unit pricing. Amazon is addressing the issue by halting billing updates and reverting to accurate data.
Neocloud Stocks Fall After Elon Musk Predicts Collapse for Over 200 Companies
Shares of neocloud companies like Nebius, CoreWeave, and Iren dropped after Elon Musk stated that over 200 neoclouds face collapse due to debt market shutdowns. This adds pressure to a sector already facing investor concerns about debt-funded expansion and the returns on AI infrastructure investments.
Google Cloud Introduces Z4M Storage-Optimized Compute Instances for AI/ML Workloads
Google Compute Engine has announced the general availability of its Z4M machine series, offering storage-optimized compute instances for AI/ML training and inference. These instances provide up to 168,000 GiB of Local SSD storage and 400 Gbps Cloud RDMA networking, reducing data-access latency by up to 60% compared to TCP.
AWS Lambda introduces Web Functions for standard HTTP web applications
AWS Lambda launched Web Functions, a new form factor enabling direct deployment of standard HTTP web applications without code rewriting. This allows developers to run port-listening servers from frameworks like Express or Next.js on Lambda, simplifying deployment and scaling for web applications and API backends.
AWS Introduces TorchNeuron for Direct PyTorch Execution on Trainium AI Chips
Amazon Web Services (AWS) has launched TorchNeuron, a new PyTorch plugin that enables direct execution of AI applications on AWS Trainium AI chips. This integration allows developers to run PyTorch models on Trainium with minimal configuration changes, supporting eager mode, `torch.compile`, and native distributed APIs, and expanding out-of-the-box support for Hugging Face text generation models.
Microsoft 365 Family and Premium plans to reduce shared OneDrive storage
Microsoft is changing its Microsoft 365 Personal, Family, Premium, and Pro subscriptions, reducing shared OneDrive storage for Family and Premium plans from 6TB to 2TB. This change, rolling out in the coming months, means subscribers using more than 2TB will need to reduce usage or pay for additional storage.
Cloudflare prototypes Zstandard compression in Pingora to save petabytes of cache storage
Cloudflare prototyped a Cache Transcoding system using Zstandard compression within its Pingora proxy, reducing cached asset sizes by two-thirds on average. This system aims to expand effective cache capacity and decrease cross-data center bandwidth usage by trading a minor CPU increase for significant storage savings.
AWS confirms permanent data loss in UAE data center after drone strikes, advises migration
Amazon Web Services (AWS) has confirmed the permanent loss of customer data in its mec1-az2 Availability Zone in the United Arab Emirates, six months after drone strikes impacted its data centers in Abu Dhabi and Bahrain. AWS is advising customers to migrate their applications to alternate regions as it cannot restore access to the affected zone and has not provided a full service restoration timeline, impacting regional cloud operations.
Google to Invest $15.1 Billion in Finnish AI Infrastructure
Google announced a 13 billion euro ($15.1 billion) investment in AI infrastructure in Finland, marking its largest single investment in Europe. This capital will be deployed through 2028 for data centers and energy projects, including a 22-year power purchase agreement with Fortum, as Finland becomes a key location for AI data centers due to available land and power.
AWS Unveils New AI and Security Features at NYC Summit
Amazon Web Services introduced new AI and security features at the NYC Summit, including Amazon Bedrock AgentCore and AWS Continuum. These updates enhance AI applications and security, offering capabilities for organizational knowledge access and proactive security measures. The announcements indicate a focus on advancing AI operations and cybersecurity strategies.
Nine PBS Regains Access to 50TB Archive After Cloud Vendor Ceases Operations
Nine PBS, a St. Louis public television station, lost access to 50TB of archival data after its cloud storage vendor, Open Source Storage (OSS), ceased operations. A judge has since ordered Iron Mountain, the data center physically housing the data, to provide Nine PBS access to the physical devices containing its 70-year archive, which includes irreplaceable historical content.
Amazon's Leo Satellite Network Surpasses 390 Operational Satellites for Initial Launch
Amazon has successfully deployed more than 390 Leo satellites, reaching an operational milestone for providing continuous internet service. This positions Amazon as a competitor to SpaceX's Starlink, despite trailing in satellite count and ahead in deployment speed. Amazon aims for commercial service availability by mid-2026, with expectations set for initial limitations in coverage and service quality.
Google Cloud's Lakehouse enables BigQuery to query Databricks Apache Iceberg tables directly
Google Cloud's borderless Lakehouse now allows BigQuery to directly query Databricks-managed Apache Iceberg tables without data duplication. This integration removes the need for complex data pipelines, reducing storage costs, latency, and fragmented security controls. It enables cross-cloud analytics by reading open-format files in Google Cloud Storage, Amazon S3, or Azure Blob Storage.
Microsoft 365 services experience widespread outage in North America
Microsoft 365 services, including Teams, SharePoint, and OneDrive, are experiencing a significant outage primarily affecting users in North America. The disruption began at 10:44 AM ET on July 23, leading to degraded functionality and access issues across multiple platforms, prompting Microsoft to advise customers to review business continuity plans.
UK's water supply insufficient for future datacentre expansion, warns trade body
Water UK has alerted MPs that the UK lacks adequate water resources for future datacentres, contradicting government growth plans. The trade body criticizes the government's failure to address water supply needs in its AI strategy, indicating potential challenges for the tech sector's expansion in water-stressed areas.
Ireland's data centers used 23% of electricity in 2025, nearing total home consumption
Data centers in Ireland consumed 23% of the country's electricity in 2025, a significant rise from previous years. This growth raises concerns over electricity demand and its impact on residents, competing closely with residential electricity use.
SpaceX applies to launch 100k new Starlink satellites for increased bandwidth
SpaceX has applied to the FCC to launch 100,000 Gen3 Starlink satellites to enhance broadband capacity. Once deployed, the network is projected to deliver multi-gigabit symmetrical speeds, positioning SpaceX to dominate the satellite internet market and meet future demand from AI devices.
Bank of England to Regulate Key Tech Firms Like Amazon and Google
The Bank of England will regulate major tech firms, starting next week, due to concerns over financial stability. Companies like Amazon, Google, Oracle, and Microsoft must ensure resilience against cyber threats and service disruptions.
Meta announces $9 billion data center in Alberta, Canada for AI infrastructure
Meta is constructing its first Canadian data center in Alberta, with an estimated cost of $9 billion. This facility will enhance Meta's AI infrastructure amid increasing competition with tech giants in cloud services.
Google Cloud C4N Instances Now Generally Available for High I/O Workloads
Google Cloud has announced the general availability of C4N, optimized for network and block storage I/O. This instance type offers up to 400 Gbps network bandwidth and 1M IOPS, addressing performance bottlenecks for enterprise applications, particularly in sectors requiring high data transfer rates.
Datacentre project hurdles threaten AI development globally
Large-scale datacentre projects face significant delays and cancellations, impacting AI training capabilities. Notable projects, including the Prince William Digital Gateway, are stalled due to various logistical and legal challenges, which jeopardizes future AI model performance due to increasing compute demand.
NVIDIA and AWS Enhance AI Production with New EC2 G7 Instances
NVIDIA and AWS have launched EC2 G7 instances powered by NVIDIA RTX PRO 4500 GPUs, enhancing AI production capabilities. These instances offer substantial performance improvements, enabling enterprises to deploy AI and data analytics workloads at scale with lower operational complexity.
Microsoft announces Azure Cobalt 200 VMs with 50% performance improvement for AI
Microsoft's Azure Cobalt 200 Arm-based VMs provide a 50% performance boost over the Cobalt 100, tailored for agentic AI workloads. This launch signals a shift in cloud architecture due to increased customer demand for compute in AI applications.
Cloudflare Halves DNS Build Delay and Cuts Memory Use by 85% for Large Zones
Cloudflare implemented a targeted build planner and a new scheduler in its DNS propagation pipeline, reducing DNS build delays by 50% and peak memory usage by 85%. These changes address the growing complexity of large DNS zones, some exceeding 20 million records, by processing zone setting changes more efficiently.
OpenAI Pilots 'Grid Protectors Program' to Offer Data Center Power to Local Grids
OpenAI announced its 'Grid Protectors Program,' piloting at a new data center in Effingham County, Georgia, to provide up to 1,000 megawatts of power to local grids during emergencies. This initiative aims to address public concerns about data center energy consumption by positioning AI campuses as community resources rather than just power consumers.
Amazon ECS Gateway Simplifies Ingress Networking for Containerized Services
Amazon introduced ECS Gateway, a new account-level resource that automates the provisioning and management of ingress networking infrastructure for containerized workloads on Amazon ECS. This solution reduces configuration overhead for platform teams and allows application teams to connect services without manual networking setup, optimizing infrastructure costs by enabling multiple services to share ALBs.
Google Illegally Clears 250,000 Trees for Finland Datacenter, Halting Construction
Google illegally cleared 250,000 trees for a datacenter site in Muhos, Finland, without completing environmental impact assessments. The Finnish Licensing and Supervision Authority has ordered a temporary halt to construction at this site and another, sparking public anger and debate over Finland's role in AI infrastructure development.