Containers stories
Developers can now isolate AI-generated code and user scripts inside Cloud Run, reducing the risk of exposing credentials or host data.
Security teams can now tighten oversight of service accounts, API keys and AI agents as machine identities outnumber staff in many enterprises.
Demand for production-ready AI agents is pushing Google to reframe Cloud Run as a platform for long-running, data-driven workloads.
UK and Ireland partners will gain wider access to Docker's container platform as Exclusive Networks expands its cloud-native software portfolio.
Enterprises and agencies handling sensitive data now have a sovereign hybrid option as DXC targets stricter compliance and AI-ready workloads.
It lets regulated and sovereign AI teams move the same model stack from a single DGX Spark to larger GPU clusters without rebuilding it.
Banks under regulatory pressure may be able to modernise databases without moving sensitive records to public cloud infrastructure.
It could simplify mixed HPC and AI estates, as pre-validated software now runs on ProLiant servers and shared systems gain tighter tenancy controls.
The update aims to give large organisations tighter control over AI workloads spread across clouds, servers and containers, with policy checks built in.
The multi-year project aims to give Telefónica Germany more control over data, resilience and AI-ready services as it modernises infrastructure.
Unsanctioned AI tools and siloed IT systems are widening risk for hospitals, as 88% say on-site infrastructure is not ready.
The new service is aimed at reducing downtime and data loss for enterprises running Kubernetes and virtual machines across hybrid HPE environments.
Research teams can now run HPC and AI workloads without rebuilding storage and cloud stacks, as CIQ adds support for major file systems.
Research centres can now keep existing Lustre and GPFS data in place as Fuzzball 4.0 adds Azure support, caching and a registry.
The platform aims to cut idle cloud spend for Kubernetes users, with DevZero saying it can shift workloads live as demand changes without restarts.
Most financial institutions now see unsanctioned AI use as a business risk, with 86% of IT executives warning of weak oversight.
Users can now route AI and HPC jobs across five clouds and on-premises through one workflow, cutting rebuilds and manual reconfiguration.
The addition gives companies a shared layer for securing and routing AI traffic as agentic systems move into production.
The update lets AI and HPC teams move workloads across five clouds and on-premises, cutting duplication and simplifying GPU access.
Cluster operators gain automated workload balancing and broader networking controls in the latest release, reducing manual intervention during maintenance.
