All news in category “Vendor and Hyperscaler Watch”

🚀 Google outlines designs for agile, fungible data centers to meet explosive AI demand, advocating modular, interoperable architectures and late-binding of facility resources. It highlights Project Deschutes liquid cooling, +/-400Vdc power proposals with Mt. Diablo side-car designs, and open efforts like Caliptra 2.0 and OCP L.O.C.K.. The post calls for community standards across power, cooling, telemetry, networking, and security to improve resilience, sustainability, and operational flexibility.

🧭 LLM-Evalkit is an open-source, lightweight application from Google that centralizes and streamlines prompt engineering using Vertex AI SDKs. It provides a no-code interface for creating, versioning, testing, and benchmarking prompts while tracking objective performance metrics. The tool promotes a dataset-driven evaluation workflow—define the task, assemble representative test cases, and score outputs against clear metrics—to replace ad-hoc iteration and subjective comparisons. Documentation and a guided console tutorial are available to help teams adopt the framework and reproduce experiments.

🔐 Google will transition in November 2025 from overwrite-based media sanitization to cryptographic erasure, using default encryption to render data unrecoverable by securely deleting encryption keys rather than overwriting drives. Recognized in NIST SP 800-88, this method is faster and better suited to modern storage technologies. Google says it will apply a layered, defense-in-depth model with independent verification, key rotations, and protections for device secrets to maintain strong safeguards.

⚙️ This post introduces the fundamentals of chaos engineering and explains why deliberately injecting controlled failures helps teams build more resilient cloud-native systems. It covers core principles — such as defining a steady-state hypothesis, limiting blast radius, replicating realistic failure modes, and automating experiments — and translates them into practical steps for experiment design, fault injection, probing, and rollback. The article recommends using Chaos Toolkit and points to Google Cloud–specific recipes to help engineers begin safely and iteratively.

🔐 Microsoft frames security culture as a company-wide movement driven by people and operationalized through the Secure Future Initiative (SFI). The company overhauled employee education—launching the Microsoft Security Academy, refreshing the Security Foundations series, and requiring three annual sessions (90 minutes total)—to address AI-enabled attacks, deepfakes, and identity threats. Leadership mandates, linked compensation, measurable training outcomes (99% completion; rising satisfaction and relevancy scores), new identity and AI guides, Deputy CISOs in engineering, and embedded DevSecOps are highlighted as evidence of measurable cultural change.

⚠️ Microsoft is investigating an ongoing incident that is preventing some customers from accessing Microsoft 365 applications. The issue has been tagged as an incident in the admin center while Redmond reviews telemetry and recent service changes to identify the root cause. Microsoft first acknowledged the problem at 05:06 AM UTC and said it continued analysis nearly four hours later to develop a fix. Impact appears limited to users served by the affected infrastructure.

🔤 Amazon QuickSight now extends font customization to data labels and axes across supported charts, complementing existing controls for titles, subtitles, legends, and table headers. Authors can set font family, pixel size, color, and styling such as bold, italics, and underline for analyses, dashboards, reports, and embedded scenarios. The update improves branding alignment and readability on large displays and is available in all supported regions.

🛡️ Varonis introduces Interceptor, an AI-native email security solution that combines multimodal AI—visual, linguistic, and behavioral models—to detect advanced phishing, BEC, and social engineering. It augments or replaces API-based filters with a phishing sandbox that pre-analyzes newly registered domains and URLs and a lightweight browser extension for multichannel protection. Integrated with the Varonis Data Security Platform, Interceptor aims to reduce false positives, accelerate detection of zero-hour threats, and stop breaches earlier in the attack chain.

🚀 Amazon EC2 High Memory U7i-12tb instances with 12TiB of DDR5 memory are now available in the Asia Pacific (Mumbai) region as of Oct 13, 2025. The u7i-12tb.224xlarge is powered by custom fourth‑generation Intel Xeon Scalable (Sapphire Rapids) processors, offers 896 vCPUs, and supports up to 100Gbps for both EBS and network throughput plus ENA Express. These instances are targeted at mission‑critical in‑memory databases such as SAP HANA, Oracle, and SQL Server, enabling higher transaction throughput, faster data loading, and shorter backup windows for large-scale deployments.

🔍 AWS Resource Explorer now provides immediate access to resource search within each AWS Region without requiring prior activation. To start searching you need, at minimum, permissions granted by the AWS Resource Explorer Read Only Access or AWS Read Only Access managed policies, and you can discover resources via the Resource Explorer console, Unified Search, or AWS CLI/SDKs. To index the full inventory, including historical backfill and automatic updates, complete Resource Explorer setup so it can create a service-linked role. You can also enable cross-Region search with a single console click or the new CreateResourceExplorerSetup API, and the feature is available at no additional cost in supported Regions.

🛠️ Amazon Web Services announced that SageMaker AI Projects can now provision custom ML project templates stored in Amazon S3. Administrators can define and manage standardized end-to-end project templates in SageMaker AI Studio so data scientists can create projects that follow organizational patterns and automated workflows. The feature is available in all AWS Regions where SageMaker AI Projects is offered.

⚠️ Microsoft says the Windows 11 Media Creation Tool (MCT) version 26100.6584 released on September 29, 2025, may close unexpectedly on Windows 10 22H2 devices without showing an error. The company is working on a fix and recommends downloading a Disk Image (ISO) for x64 systems as a temporary workaround. Microsoft also notes the MCT is not supported on Windows 10 ARM64 machines, following earlier ARM64 compatibility problems after the Windows 11 25H2 rollout.

🔐 Amazon Bedrock AgentCore is now generally available, offering an agentic platform to build, deploy, and operate agents securely at scale without infrastructure management. It adds VPC support, extended eight-hour runtimes, session isolation, and initial Agent-to-Agent (A2A) protocol support. Identity, Gateway, and Observability features provide secure authorization, tool discovery, and CloudWatch/OTEL monitoring.

🔐 Apple has increased the top award in its Apple Security Bounty program to $2m for exploit chains that emulate sophisticated mercenary spyware. Bonuses for Lockdown Mode bypasses and vulnerabilities found in beta software can more than double that payout, potentially exceeding $5m. Apple also raised many category rewards — including $100,000 for a Gatekeeper bypass and $1m for broad unauthorized iCloud access — and introduced a Target Flags initiative to speed and standardize exploitability demonstrations.

📣 AWS Config now supports three additional resource types—AWS::ApiGatewayV2::Integration, AWS::CloudTrail::EventDataStore, and AWS::Config::StoredQuery—providing broader visibility across AWS environments. If you have recording enabled for all resource types, AWS Config will automatically begin tracking these new types. They are available for use in Config rules and Config aggregators in all Regions where the resources exist. This expansion enhances your ability to discover, assess, audit, and remediate a wider range of resources.

🔍 Amazon CloudWatch is generally available with Generative AI Observability, providing end-to-end telemetry for AI applications and AgentCore-managed agents. It expands monitoring beyond model runtime to include Built-in Tools, Gateways, Memory, and Identity, surfacing latency, token usage, errors, and performance across components. The capability integrates with orchestration frameworks like LangChain, LangGraph, and Strands Agents, and works with existing CloudWatch features and pricing for underlying telemetry.

🚀 Amazon ElastiCache now offers vector search generally available with Valkey 8.2, enabling indexing, searching, and updating billions of high-dimensional embeddings from providers such as Amazon Bedrock, Amazon SageMaker, Anthropic, and OpenAI with microsecond latency and up to 99% recall. Key use cases include semantic caching for LLMs, multi-turn conversational agents, and RAG-enabled agentic systems to reduce latency and cost. Vector search runs on node-based clusters in all AWS Regions at no additional cost, and existing Valkey or Redis OSS clusters can be upgraded to Valkey 8.2 with no downtime.

📢 AWS is updating the availability status of a range of services and features across three lifecycle categories: moving to Maintenance, entering Sunset, and reaching End of Support. Services moving to maintenance will stop accepting new customers on Nov 7, 2025, while current customers can continue using them as they evaluate alternatives. Several services, including Amazon FinSpace, AWS IoT Greengrass v1, and AWS Proton, are entering sunset with documented timelines (typically ~12 months). AWS Mainframe Modernization App Testing has reached end of support as of Oct 7, 2025. AWS provides migration guides and support resources to help customers transition.

⚠ Microsoft warned that devices running Windows 11 23H2 Home and Pro editions will stop receiving security updates after November 11, 2025. The November 2025 monthly security update will be the final update for those editions. Users should upgrade to Windows 11 24H2 or later to remain protected; note that some PCs may be prevented from upgrading by a safeguard for SenseShield code-obfuscation drivers.

🔒 Apple has raised its top bounty for iOS zero‑click system‑level remote code execution from $1 million to $2 million, with additional bonuses for Lockdown Mode bypasses and beta‑stage reports that can push awards above $5 million. The change coincides with the rollout of Memory Integrity Enforcement in A19/A19 Pro chips, which leverages Arm's MTE/EMTE to harden memory safety. Apple will also provide 1,000 iPhone 17 devices to civil society members at risk.