All news in category “Vendor and Hyperscaler Watch”

🧊 Google announced preview read/write interoperability between BigQuery and Iceberg-compatible engines via the Google-managed Iceberg REST Catalog. The capability lets BigQuery, Trino, Spark, Flink and others create, update, and query a single Iceberg table type while enforcing unified governance and table-level access controls. Customers can offload compaction and garbage collection to BigLake to reduce small-file and metadata bloat and improve query performance.

🔧 Palo Alto Networks announces the general availability of the ADEM Universal Agent, a hardware-agnostic telemetry agent for Prisma Access designed to deliver consistent, high-fidelity data from branch and edge sites. The agent can run on VMs or containers, enabling synthetic testing, hop-by-hop path analysis, and overlay/underlay visibility regardless of on-prem hardware. By consolidating disparate telemetry into a unified data engine, the agent reduces blind spots and accelerates root-cause identification to support automated, machine-speed operations.

🔍 Orchid Security warns that modern IAM estates harbor extensive "identity dark matter," with roughly 46% of identity activity operating outside centralized visibility. The article positions Gartner's Identity Visibility and Intelligence Platform (IVIP) as a necessary observability layer that unifies telemetry from managed and unmanaged systems, applies AI to infer intent and risky behavior, and enables automated remediation to reduce exposure.

🔧 This post walks through deploying and serving a large model on Google Kubernetes Engine using managed DRANET and NVIDIA B200 GPUs. It explains how RDMA networking is provisioned as an isolated regional VPC for low-latency GPU-to-GPU communication and how to provision A4 nodes and reservations for RoCEv2-capable accelerators. The author provides example gcloud and kubectl commands to create the cluster, a GPU node pool with DRA labels, a ResourceClaimTemplate for mrdma workloads, and steps to serve a DeepSeek model privately via GKE Inference Gateway and a regional internal Application Load Balancer.

🔒 Kaspersky highlights that security management consoles themselves expand an organization’s attack surface and therefore must be hardened. Kaspersky Security Center Linux 16.1 adopts a secure-by-default model by enabling two-factor authentication for all console access and removing the global option to disable it. Administrators are required to ensure 2FA is configured for users who access the Web Console or use OpenAPI automation before upgrading. Kaspersky also publishes a structured hardening checklist to audit roles and privileges, restrict network access, strengthen encryption, protect APIs, and ensure comprehensive logging and auditing.

🚀 Oracle Database@AWS is now generally available in five additional AWS Regions — EU-West-1 (Dublin), EU-West-2 (London), AP-South-1 (Mumbai), AP-South-2 (Hyderabad), and AP-Northeast-2 (Seoul) — expanding coverage to twelve Regions. The service enables AWS customers to access OCI-managed Oracle Exadata systems from within AWS data centers, supporting in-region data residency and migrations of on-prem Exadata and RAC workloads. Dublin, Mumbai, and Hyderabad offer two Availability Zones while London and Seoul currently provide one; CA-Central-1 and AP-Southeast-2 now support two AZs for enhanced production availability. To consume the service, request a private offer from Oracle via the AWS Marketplace and provision databases through the AWS Management Console.

🚀 AWS Lambda response streaming is now available in all commercial AWS Regions, enabling the InvokeWithResponseStream API to progressively stream response payloads as they are produced. This reduces time-to-first-byte (TTFB) for latency-sensitive workloads such as LLM-based, web, and mobile applications by allowing partial responses up to a default 200 MB. Response streaming is supported via AWS SDKs, Amazon API Gateway REST APIs, Node.js managed runtimes, and custom runtimes; note that additional network transfer charges apply for the bytes streamed out over the initial 6 MB.

🔒 Cloudflare is accelerating its post-quantum roadmap and now targets 2029 to achieve full post-quantum security, explicitly including post-quantum authentication. The company already enabled post-quantum encryption for the majority of human traffic to mitigate harvest-now/decrypt-later risks, but new algorithmic and hardware advances (notably Google’s reported speedups and Oratomic’s neutral-atom estimates) make authentication the urgent priority. Cloudflare will enable PQ defaults for customers at no extra cost.

🛡️ At Nutanix .NEXT 2026, Palo Alto Networks highlighted an expanded integration delivering native, automated security across Nutanix environments and was named Nutanix 2026 Global Security Partner of the Year. The partnership extends Layer‑7 protection via VM‑Series virtual firewalls, consistent hybrid cloud policies for Nutanix Cloud Clusters (NC2), and Panorama-driven automation. A forthcoming integration embeds Prisma AIRS into Nutanix Enterprise AI (NAI) to enforce AI Model Security, continuous AI Red Teaming, and unified visibility so only validated models reach production.

🛡️ At AWS, decades of scaled security operations combine with new AI collaborations to proactively harden critical systems. Through Project Glasswing and Anthropic’s Claude Mythos Preview, AWS runs continuous AI-driven code reviews and provides gated research previews via Amazon Bedrock. Complementary offerings include AWS Security Agent for autonomous penetration testing and Bedrock guardrails and Automated Reasoning to enforce enterprise controls and reduce risk.

🔒 Anthropic’s newest and most capable model, Claude Mythos Preview, is available in Private Preview to a select group of Google Cloud customers through Project Glasswing. Its placement on Vertex AI provides enterprises access to a frontier model integrated with Google Cloud’s tools to build, scale, and govern AI applications and agents. The announcement emphasizes high performance across use cases and a renewed focus on reducing cybersecurity risk in enterprise deployments.

🚀 Amazon Lightsail is now available in the Asia Pacific (Malaysia) Region, bringing Lightsail's simplified cloud compute and networking to customers in Malaysia and neighboring countries. The launch offers lower latency, improved performance and helps meet local data residency requirements. Customers gain access to instances (general purpose, compute- and memory-optimized), managed databases, containers, load balancers and predictable pricing via the Lightsail Console, AWS CLI and SDKs.

🔒 Google Cloud's Cloud NGFW Enterprise now supports domain and SNI-based URL filtering with limited wildcard matching to shift enforcement to the application layer. The URL filtering service inspects HTTP payloads and SNI headers to enable granular egress policies and block malicious domains without requiring full TLS decryption. This reduces the operational burden of tracking dynamic IPs and helps prevent bypass techniques such as SNI spoofing while preserving end-to-end encryption and compliance.

⚙️ Amazon SageMaker Unified Studio now supports Serverless Workflows in Identity Center domains, allowing customers to orchestrate data-processing tasks with Apache Airflow (via Managed Workflows for Apache Airflow) without provisioning Airflow infrastructure. Serverless Workflows auto-provision compute during runs and release it afterward, so you pay only for actual run time. Each workflow runs with its own execution role and isolated worker to ensure workflow-level security and prevent cross-workflow interference. The Visual Workflow experience supports around 200 operators and built-in integrations with services such as Amazon S3, Amazon Redshift, Amazon EMR, AWS Glue, and Amazon SageMaker AI.

🔒 Amazon Bedrock now offers Claude Mythos Preview in a gated research preview as part of Project Glasswing. Anthropic's most advanced model to date demonstrates state-of-the-art capabilities across cybersecurity, software coding, and complex reasoning, identifying sophisticated vulnerabilities and showing exploitability in large codebases with less manual guidance. Access is limited to an allow-list in US East (N. Virginia) through Bedrock; AWS account teams will contact approved organizations.

🏠 Rightmove migrated from siloed on-premises databases to Google Cloud to build a unified analytics and AI platform it calls the data hive. Using BigQuery, Vertex AI, and Looker, the company extracts metadata from listings and images to deliver personalized search, agent-assist messaging, and an Automated Valuation Model. The hub-and-spoke architecture centralizes governance while enabling business units to run tailored forecasting and ML use cases. Around 300 staff now use the platform to convert data into operational and commercial value.

🎵 This guide outlines best practices for prompting Lyria 3 and Lyria 3 Pro, Google’s music generation models that deliver granular control over vocals, instrumentation, arrangement, and timing. It highlights technical details—track lengths from rapid 30-second prototypes to three‑minute compositions, multi‑vocal support in eight languages, timed-lyrics and tempo conditioning—and includes a concise prompting framework. The post also covers advanced workflows such as timestamped segment instructions and multimodal generation using images or PDFs, plus integration paths through Vertex AI and the Gen AI SDK.

🎵 Google has made Lyria 3 and Lyria 3 Pro available on Vertex AI in public preview, bringing high-fidelity music generation to the Vertex AI API and Media Studio. Lyria 3 Pro composes studio-quality tracks up to three minutes with structural elements (intros, verses, choruses, bridges), while Lyria 3 produces 30-second tracks for rapid prototyping. Both models accept multi-modal inputs (text or images), support vocal generation with timed lyrics or user-provided lyrics, and can produce purely instrumental pieces. Outputs are embedded with SynthID watermarking and filtered for policy and IP compliance.

📁 S3 Files provides a shared, high-performance file system that lets any AWS compute resource access data directly in Amazon S3 with full file-system semantics and low-latency performance, without moving objects out of S3. Built on Amazon EFS, it maintains a live view of bucket objects and translates file operations into efficient S3 requests so applications and agents run unchanged. It caches active data for fast reads, delivers multi-terabytes-per-second aggregate throughput, and is generally available in 34 AWS Regions.

🔍 The article, by Sila Ozeren Hacioglu of Picus Security, describes the 'PoC Cliff' where automated pentesting delivers strong initial results but rapidly dwindles after a few executions as its deterministic, chained approach exhausts favored attack paths. It contrasts that model with Breach and Attack Simulation (BAS), which runs thousands of independent, atomic tests to validate whether defenses actually detect and block techniques. The piece identifies six critical validation surfaces often left dark and gives three diagnostic vendor questions to close the gap.