Cloud Platforms EKS Auto Mode Resizes Pod Requests In Place From Observed Usage EKS Auto Mode resource optimization rewrites a Pod's CPU and memory requests from measured usage, in place and without touching the template. Here is how it behaves.
Cloud Platforms Cloudflare Turns a Spending Cap Into a Runtime Decision "Spend Management pauses Workers and Durable Objects at a dollar limit, and the engineering story sits in the path from usage event to runtime rejection: two metering stages, an account-part
Cloud Platforms Microsoft Ends Unmanaged Developer Environments in Power Platform "Microsoft stops creating unmanaged developer environments on 15 November 2026, suspends the existing ones on 15 December and starts deleting them on 15 January 2027, and the conversion depe
Cloud Platforms Compute Engine Adds Immutable Date Versions Beside the v1 API "Compute Engine API multi-versioning is generally available: quarterly stable versions such as 2026-09-01 freeze schemas and defaults, previews live for a year, and the version travels in a
Cloud Platforms Google Cloud's Z4M Puts 168,000 GiB of Local SSD Behind AI Checkpoints "Google's Z4M storage-optimized series reached general availability on 6 November 2026 with up to 168,000 GiB of local SSD, 875 GiB per vCPU, 400 Gbps Cloud RDMA, and a maintenance rule that
Cloud Platforms Google Cloud Spend Caps Reach General Availability With an API and Live Costs Google Cloud spend caps are generally available with an API on the Budgets resource, live cost reports that explain a trigger, and an automatic monthly reset, while Provisioned Throughput ke
Cloud Platforms A Cost Comparison Method for Leaving the Public Cloud A cloud exit is decided by arithmetic that starts in twelve months of invoices. This method splits every line into storage, transfer and compute, prices each bucket at published rates, adds
Cloud Platforms Scaleway Adds GPU Zones in Two Countries on Two Accelerator Families Scaleway announced on 13 October 2026 that it is adding two new clusters, an AMD Instinct MI355X in Stockholm and an NVIDIA B300 in Copenhagen, alongside its GPU Zone in Paris, plu
Cloud Platforms Cloudflare Workers AI vs Amazon Bedrock: Edge Inference or Managed Enterprise Model Access? "Workers AI puts 50+ open models on Cloudflare's edge GPUs behind pay-per-use pricing. Bedrock puts 100+ hosted models — including frontier labs — behind enterprise controls. They are not co
Cloud Platforms The Cloud Is Still Shared Hardware: What Cloudflare's Spectre Reassessment Changes Cloudflare's new Workers research leaked a JWT at up to 12 bits per second with more than 99% accuracy in production. The important lesson is not that edge isolation failed; it is that isola