All news in category “Vendor and Hyperscaler Watch”

🖥️ Amazon Redshift's ODBC 2.x driver now provides native support for Apple macOS, enabling developers and analysts to connect to Redshift clusters from macOS using third-party SQL clients and applications. The native driver exposes Redshift capabilities such as data sharing write and Amazon IAM Identity Center integration that are available only through Redshift drivers. This update streamlines integration with ETL and BI tools on macOS, and Amazon recommends upgrading to the latest ODBC 2.x driver and consulting the driver documentation for installation and system requirements.

🔁 AWS Glue zero-ETL now supports ongoing replication from self-managed Oracle, SQL Server, MySQL, and PostgreSQL databases into Amazon Redshift in seven additional AWS Regions. The no-code integration automatically creates and manages an ongoing replication pipeline, removing much of the setup and configuration complexity. This capability reduces operational burden and can save weeks of engineering effort for teams migrating or consolidating data into Redshift. Newly added regions include Hong Kong, Tokyo, Singapore, Sydney, London, São Paulo, and US (Virginia).

🔖 AWS now supports using AWS Organizations account tags across its cost management products. Account tags applied at the organization level automatically flow to all metered usage in tagged accounts, removing the need to maintain separate account lists in Cost Explorer, CUR 2.0/FOCUS, AWS Budgets, and Cost Categories. This also enables allocation for untaggable items like refunds, credits, and certain service charges. To enable, add tags in the AWS Organizations console and activate them on the Cost Allocation Tags page in Billing. The feature is generally available in all Regions except GovCloud (US) and China (Beijing/Ningxia).

🔁 Amazon Elastic Container Registry (ECR) can now automatically create repositories when images are pushed, removing the need to pre-create repositories before a push. Repository creation follows organization-defined repository creation templates, enabling consistent naming and default settings at creation. The create-on-push capability is available in all AWS commercial and AWS GovCloud (US) Regions.

🔁 Amazon Timestream for InfluxDB now provides a restart API for both InfluxDB 2 and InfluxDB 3, enabling customers to trigger instance restarts via the AWS Management Console, API, or CLI. The capability supports resilience testing and direct remediation of health-related issues without requiring support intervention. It is designed to give DevOps teams greater operational flexibility for mission-critical time-series workloads and is available in all Regions where the service is offered.

🔧 Amazon Bedrock Data Automation now offers blueprint instruction optimization to improve custom field extraction accuracy using just a few example document assets with ground truth labels. The feature analyzes differences between expected values and Data Automation inferences, then refines the natural-language instructions in your blueprint to boost extraction performance without model training or fine-tuning. It produces evaluation metrics including exact match rates and F1 scores and is available in all Regions where Bedrock Data Automation is supported.

🖥️ Amazon WorkSpaces Applications now supports Microsoft Windows Server 2025, enabling customers to launch streaming instances powered by the newest server OS. The update delivers improved security, performance, and modern platform capabilities, and allows administrators to provide a Windows 11-style desktop experience to end users. Images are available as AWS public images or can be customized with Image Builder, and support is generally available in all Regions where the service is offered.

🔬 Stochastic rounding restores tiny gradient updates that deterministic low-precision formats would otherwise zero out, enabling stable training in FP8 and 4‑bit regimes. Frameworks such as JAX and the Qwix quantization toolkit apply SR on Google Cloud accelerators—TPU MXUs and NVIDIA Blackwell A4X VMs—to prevent vanishing updates. The approach trades deterministic bias for unbiased noise, often acting as implicit regularization and preserving model convergence while boosting efficiency.

🚀 These hands-on labs guide developers through three Google Cloud deployment options to move AI agents from local prototypes to production. The Vertex AI Agent Engine offers a fully managed, Python-optimized runtime that handles execution, memory, and tool invocation. Cloud Run provides a serverless container experience with autoscaling and language flexibility, while GKE delivers orchestrated control for microservice deployments.

🔁 Amazon EC2 now accepts Availability Zone ID (AZ ID) as a parameter in its APIs, enabling creation and management of instances, volumes, subnets and other resources using consistent, static zone identifiers. AZ IDs map to the same physical location across all AWS accounts, removing the need to reconcile per-account AZ names. This capability covers launch templates, fleets, reserved instances, volumes, capacity reservations, VPC endpoints, network interfaces, fast snapshot restore, instance connect and more, and is available in all regions including China and AWS GovCloud (US).

🔬 Google is partnering with the DOE's Genesis Mission to accelerate federally funded scientific discovery by combining high-performance computing, experimental facilities, and AI. Gemini for Government and Google DeepMind tools offer multimodal reasoning, agentic workflows, and an AI co‑scientist to speed hypothesis development. Google Cloud provides the secure, accredited infrastructure for multi‑lab deployments.

🔍 Google Cloud describes how the AlloyDB AI natural language API translates user questions into SQL and how to tune it for near‑perfect accuracy in enterprise applications. The post outlines a hill‑climbing workflow that improves results by adding descriptive (table and column) and prescriptive (templates, facets) context, plus an automated value index for private terms. It highlights capabilities for business relevance, explainability, and verified results, and explains agent integration options such as the MCP Toolbox and Gemini Enterprise.

🔒 Microsoft’s new e-book, "3 reasons point solutions are holding you back", argues that fragmented security tools increase costs, slow investigations, and limit AI effectiveness. It advocates a unified, AI-ready security platform that consolidates telemetry, analytics, and automation across detection, response, exposure management, and cloud security. Learn how Microsoft Defender, Microsoft Sentinel, and Microsoft Security Copilot combine to improve MTTR, predictive defense, and operational efficiency.

🚀 Amazon Aurora PostgreSQL-Compatible Edition now supports PostgreSQL community releases 17.7, 16.11, 15.15, 14.20, and 13.23. The update combines upstream community bug fixes and product improvements with Aurora-specific optimizations, including faster Blue/Green deployment switchovers and enhancements to Query Plan Management (QPM). These engine versions are available in all commercial AWS Regions and AWS GovCloud (US), and can be deployed or used to upgrade existing clusters via the Amazon RDS console.

🛡️ Google Cloud has integrated Cloud API Registry into Vertex AI Agent Builder, giving administrators centralized governance over agent tools and a curated catalog developers can access via a new ApiRegistry ADK object. The update broadens ADK support — including Gemini 3 Pro/Flash and TypeScript — and improves state management, interactions, and recovery. An early A2UI toolkit and Interactions API support aim to simplify multimodal I/O and shared UI components. Agent Engine features such as Sessions and Memory Bank are GA, regional availability is expanding, and several pricing adjustments take effect in December 2025 and January 2026.

🐧 Amazon WorkSpaces Applications now supports Ubuntu Pro 24.04 LTS on Elastic fleet, enabling Independent Software Vendors and central IT to stream Ubuntu desktop applications from an AWS-managed, serverless pool. Elastic fleet removes the need for capacity planning, scaling policies, or image creation. Administrators can enable the feature via the WorkSpaces Applications console and choose supported AWS Regions; pricing is pay-as-you-go.

🚀 Amazon Web Services has launched the compute-optimized EC2 C8a instances in the Europe (Spain) region. Powered by 5th Gen AMD EPYC processors running up to 4.5 GHz, C8a offers up to 30% higher performance and up to 19% better price-performance versus C7a, plus 33% greater memory bandwidth. Available in 12 sizes (including two bare-metal options), they target high-performance, latency-sensitive workloads and support Savings Plans, On-Demand, and Spot purchasing.

🔔 Amazon ECS Managed Instances now supports Amazon EC2 Spot Instances, enabling customers to run fault-tolerant container workloads on spare capacity at discounts up to 90% versus On‑Demand. The fully managed compute option continues to provision, configure, and operate EC2 instances in your account while AWS handles scaling and placement optimization. You set task requirements and choose capacityOptionType as spot or on-demand in the Managed Instances capacity provider. Spot support is available in all Regions where Managed Instances is offered; management fees apply in addition to Spot instance costs.

🌐 Amazon MSK Connect now supports dual-stack connectivity (IPv4 and IPv6) for new connectors on MSK Connect, allowing customers to create connectors that use both protocols. Connectors can be created with dual-stack via the Amazon MSK Console, AWS CLI, SDKs, or CloudFormation by setting the Network Type parameter. New connectors default to IPv4-only unless explicitly configured for dual-stack, and existing connectors remain IPv4 and must be deleted and recreated to change. The feature is available in all Regions where MSK Connect is offered and incurs no additional charge.

🌐 Amazon WorkSpaces now supports IPv6 for WorkSpaces domains and external endpoints, enabling IPv4/IPv6 dual‑stack connectivity from compatible clients while excluding SAML authentication over IPv6. Dual‑stack reduces the need for address translation hardware, simplifies IP management, and supports PrivateLink VPC endpoints over IPv6 for private access. The feature is available in all AWS Regions, including GovCloud, at no additional cost; administrators must use the latest WorkSpaces clients to enable IPv6.