All news with #google cloud tag

⚡ The Ultra Low Latency (ULL) Solution and new U4 machine family are now GA, delivering deterministic, low-jitter compute and hardware-accelerated multicast networking for high-frequency trading workflows in Google Cloud. The solution combines bare metal and VM options, precision timing, 24/7 packet capture, and isolated multicast fabrics to match co-location performance while providing cloud elasticity and observability. Available in select private regions, it targets exchanges, market participants, and trading service providers.

🔍 This article explains how AlloyDB AI streamlines hybrid search for modern AI and RAG applications by combining vector search and full-text search into a single SQL function using Reciprocal Rank Fusion (RRF). It highlights new FTS capabilities including the RUM extension for positional indexing and the BM25 index for industry-standard ranking. The post also describes an external search Foreign Data Wrapper (FDW) to integrate Elasticsearch, OpenSearch, and Solr while maintaining a unified SQL interface.

🚀 The Lakehouse runtime catalog is a fully serverless, Spanner-backed implementation of the Apache Iceberg REST catalog designed to provide high availability, strong consistency, and horizontal scale for metadata management. It decouples metadata discovery from compute engines to enable multi-engine interoperability and supports credential vending, governance integration, and bi-directional federation. The catalog aims to reduce operational overhead and support agent-scale workloads with enterprise-grade features.

🚀 Google Cloud announces Google Cloud Modernize, an end-to-end portfolio that consolidates migration and modernization tools to accelerate enterprise transformation with AI. The offering centers on Modernization Hub, an in-console experience for analyzing code and mapping dependencies for Java, .NET, and mainframe apps. New Gemini-powered capabilities in Migration Center provide rapid TCO estimates and interactive cost modeling. Purpose-built compute, VMware support, and agentic migration tools (EKS-to-GKE) help organizations modernize infrastructure and application estates with enterprise-grade controls.

🔧 AI21 Labs adopted Google Cloud AI Hypercomputer and Kueue on GKE to run foundation models like the Jamba family at scale. They pooled thousands of A3 and A3 Ultra GPU instances into a shared GKE cluster to maximize utilization and replaced manual capacity negotiation in Slack with automated scheduling. The change reduced high-priority job wait times from 72 hours to 12, cut manual scheduling interventions from 20 per week to zero, and lowered fragmentation.

🧠 This article outlines a two-tier memory architecture for enterprise AI agents that pairs Memorystore for Valkey as a short-term session buffer with AlloyDB AI for long-term persistent memory. It explains how the design reduces token consumption, improves latency, and preserves transactional integrity while supporting hybrid retrieval, in-database embeddings, and generative functions. The post includes benchmark results showing significant token and latency savings and summarizes implementation patterns and SQL functions used to build the system.

🚀 Spanner queues are now generally available, embedding native transactional messaging directly within Spanner to support reliable agentic execution. Creating a message is a single write within the same ACID transaction that updates an agent's state, enabling atomic decide-and-act semantics, scheduled deliveries, streaming SQL pulls, and exactly-once processing guarantees. The feature simplifies multi-agent orchestration, timeouts, human approvals, and observability using standard GoogleSQL.

🔒 Google Cloud now enables end-to-end checksumming by default across all Cloud Storage SDKs to strengthen data integrity. The SDKs compute and pass checksums during uploads if the application does not, and verify checksums on downloads and range reads via gRPC. Cloud Storage maintains checksums at multiple internal layers—chunks, shards, blocks—ensuring a continuous chain-of-custody from application to disk. Users are advised to update to the latest SDK versions to benefit from these protections.

🔍 PayPal migrated its legacy, on-premise Hadoop analytics platform to Google’s Managed Service for Apache Spark to streamline operations and accelerate insights. The move delivered rapid provisioning, elastic scaling, and native integration with Google Cloud Storage and BigQuery, reducing operational overhead and data silos. Results included 25% faster processing, 30% better SLA adherence, lower costs, and more engineering time for innovation.

🚀 The Google Cloud team has released google-cloud-swift, an official server-side SDK built for Swift 6.2+ that leverages Swift NIO, HTTP/2, gRPC, and compile-time concurrency safety. Designed for high-throughput microservices and DevOps tooling, the SDK supports over 100 Google Cloud services and works on macOS and Linux. It enables ARC-based deterministic memory management, async/await ergonomics, and integrates with Application Default Credentials and Workload Identity Federation for secure authentication.

🛠️ The Google Cloud CLI remote MCP server is now available in public preview, enabling AI agents to run gcloud and bq commands from a secure, network-isolated execution sandbox. This managed server removes the need to install CLI binaries locally, supports hosted agent platforms, and enforces enterprise-grade controls including zero ambient credentials, IAM-based permissions, Model Armor screening, and Cloud Audit Logging. Agents connect via the MCP standard and authenticate through Agent Identity or OAuth 2.0.

🛠️ Today Google Cloud announces the general availability of Data Agent Kit, a free set of Model Context Protocol (MCP) tools and Google-authored agent skills that let coding agents access and act on Google Cloud data products. The GA release adds support for BigQuery Graph, Bigtable, and Managed Service for Apache Spark in Lakehouse workflows, plus numerous quality-of-life and performance improvements. The kit integrates with IDEs, Antigravity, Claude Code, Codex, Cloud Shell, and Cloud Workstations, enabling agents to inspect schemas, run queries, read job logs, and orchestrate pipelines using customers' IAM permissions.

🔒 This post compares two reference networking architectures for AI inference model serving: one tailored for Google Kubernetes Engine (GKE) and one for mixed or alternative backends. It explains a common control-plane pattern using Private Service Connect, optional Apigee, and Model Armor as a centralized entry point for secure, private inference calls. The GKE design adds a specialized GKE Inference Gateway, inference pools, and replica sets for GPU/TPU workloads. The multi-backend design uses a regional internal Application Load Balancer, a Cloud Run payload processor service extension to inject model headers, and Network Endpoint Groups to route to heterogeneous backends.

🚀 Google Cloud announces GKE Agent Sandbox and the Agent Sandbox RL orchestration SDK, now generally available, to address infrastructure bottlenecks in large-scale agentic reinforcement learning. The solution integrates SandboxWarmPool with GKE Image Streaming to eliminate cold-starts, reduce GPU idle time, and support thousands of large images with low TTFC. Native integrations with popular RL tools and an async Python SDK simplify orchestration and warm-pooling strategies for researchers.

🔒 Google Cloud announced an expanded catalog of partner-built security agents and integrations for Gemini Enterprise, enabling firms to orchestrate multi-step, AI-powered security workflows from a single interface. The partners provide protections spanning deception, identity containment, runtime LLM safety, data discovery, application and cloud risk assessments, and SOC orchestration. These agents let organizations apply context-aware defenses across identity, endpoint, cloud, and agentic AI workloads while reducing containment time and operational friction. The ecosystem supports both vendor agents and protections for agentic workloads to help secure AI-driven operations.

🚀 Google Cloud announces GA for the Storage-optimized Z4D machine series across Compute Engine VMs and bare-metal instances. Built on 5th Gen AMD EPYC (Turin) and Titanium SSDs, Z4D delivers up to 84,000 GiB local SSD, up to 384 vCPUs, and 3,072 GiB memory to accelerate IO-intensive workloads by up to 40% over Z3. Z4D supports Hyperdisk variants for scalable network-attached storage and offers bare-metal support for Nutanix Cloud Clusters (NC2) in preview. Select regions have VMs available now; bare-metal is in preview.

🚀 Memorystore for Valkey 9.1 is now generally available on Google Cloud, delivering up to 3x queries per second at microsecond latencies compared to Memorystore for Redis Cluster. Valkey 9.1 introduces a lock-free multi-queue I/O architecture, a two-phase dynamic thread scaling engine, and several new commands (HGETDEL, MSETEX, HSETEX) plus topology-aware CLUSTERSCAN and database-level ACLs for improved security and observability. The release also expands node sizes and provides a managed migration workflow to simplify moving from self-managed Redis/Valkey to Google Cloud Memorystore.

🔧 This guide explains Google Cloud's managed Reinforcement Learning Fine-Tuning (RLFT) service for adapting Gemini models using a reward signal you define instead of labeled answers. It covers when to choose RLFT versus supervised fine-tuning (SFT), example use cases (NPC dialogue, structured extraction, moderation, code execution, slide generation), and the practical artifacts you must supply: a dataset and a robust reward function. The service manages infrastructure and model internals while you iterate on reward and validation.

📣 Google Cloud announces the GA release of Storage Intelligence advisor and expanded storage batch operations to help teams detect anomalies and remediate them at scale. Advisor provides out-of-the-box findings and baselines daily activity to surface spikes, errors, cross-region egress, and growth trends without heavy engineering. Batch operations let you execute large-scale changes (transitions, deletes, retention updates) across millions of objects with serverless execution, dry-run validation, and advanced filters.

🚀 Google Cloud was named a Leader in the 2026 Gartner® Magic Quadrant™ for Container Management, ranking highest for Ability to Execute. The accompanying Gartner Critical Capabilities report placed Google Cloud first across all evaluated use cases, including AI training and inference. Google highlights recent GKE and Cloud Run innovations—like predictive latency boosts, rapid startup times, serverless GPU scale-to-zero, and Agent Substrate/Sandbox—to support enterprise and AI workloads at scale. The post invites users to try the open-source Agent projects and explore Cloud Run and GKE enhancements.