Cloud

Cloud Platforms

Azure 106AWS 112GCP 109

327 articles · latest first

Gemini 4 Argon on Google Cloud: model ID gemini-4-argon, pricing, and Cloud Run sandboxes

Gemini 4 Argon is on Google Cloud via the Gemini API; with Cloud Run sandboxes, teams now must rethink per-request isolation, routing, and cost controls.

Oct 17, 2026·3mgoogle-cloudgemini-4-argon

Azure

Azure Maia 300 architecture: rack-local accelerator platform for agentic inference

Azure Maia 300 makes rack-local accelerators, networking, and power a single platform, forcing platform teams to adopt topology-aware scheduling and GitOps.

Oct 17, 2026·3mazuremaia-300

AWS

AWS Lambda Web Functions: multi-concurrent HTTP runtimes with automated global placement

Lambda Web Functions let a single serverless instance host web frameworks, serve concurrent HTTP requests, and be globally placed—forcing teams to rethink security.

Oct 17, 2026·3mlambda-web-functionsamazon-bedrock

GCP

Google Cloud ADK: five-language Agent Development Kit with A2A and Cloud Run/GKE scaffolding

Google Cloud ADK supports Python, Go, TypeScript, Java and Kotlin, implements A2A, and scaffolds deploys to Cloud Run, Vertex/Gemini, or GKE for production use.

Oct 16, 2026·3mgoogle-cloudagent-development-kit

Azure

Azure Maia 300: rack-local accelerator platform for agentic inference

Microsoft detailed the third-generation Azure Maia 300: a rack-local accelerator platform for low-latency, stateful agent inference and an open HSM push today.

Oct 15, 2026·3mazure-maia-300rack-local-accelerators

AWS

AWS Lambda Web Functions: run HTTP frameworks with multi-concurrency and global distribution

Lambda Web Functions run HTTP frameworks with concurrency and global distribution, forcing teams to rethink runtime isolation, operational telemetry, and IAM.

Oct 15, 2026·3maws-lambdaserverless

GCP

Google Cloud adds agent-native APIs and Cloud Run/GKE primitives for Gemini and Vertex AI

Google Cloud exposes agent-native APIs and Cloud Run/GKE execution primitives for Gemini models and Vertex AI, enabling sandboxed, long‑running agent workflows.

Oct 14, 2026·3mgcpgke

Azure

Azure Maia 300: rack-local accelerator platform for low-latency agent inference

Azure Maia 300 is Microsoft’s rack-local AI accelerator platform for agentic inference, emphasizing rack topology, fabric latency, and hardware attestation.

Oct 14, 2026·3mazureazure-maia-300

AWS

AWS Lambda Web Functions (multi-concurrency): runtime, observability, and security implications

How AWS Lambda Web Functions' multi-concurrency and global distribution change runtime, observability, and security assumptions for serverless web apps today.

Oct 14, 2026·3maws-lambdaserverless

AWS

AWS Lambda Web Functions: run web frameworks with multi-concurrent, regionally distributed runtimes

Lambda Web Functions let you run web apps with multi-concurrent runtimes and regional/edge deployment. Platform teams must treat them as long-lived services.

Oct 13, 2026·3maws-lambdaserverless

GCP

GKE 1.37: microVM sandboxes (Kata Containers + Cloud Hypervisor) reach GA

GKE 1.37 promotes microVM sandboxes (Kata Containers + Cloud Hypervisor) to GA, giving platform teams supported hardware-isolated pods for multi-tenant AI.

Oct 13, 2026·3mgkekata-containers

Azure

Azure Maia 300 architecture: co-designed platform for rack-local agent inference

Azure Maia 300 is a co-designed stack of accelerators, runtime, rack networking, and ops for rack-local agent inference, forcing placement and network changes.

Oct 13, 2026·3mazure-maia-300microsoft-foundry

315 more