All news with #google cloud tag

🧬 Escalante leverages JAX's functional, composable design to combine many predictive models into a single differentiable objective for protein engineering. By translating models (including AlphaFold and Boltz-2) into a JAX-native stack and composing them serially or linearly, they compute gradients with respect to input sequences and evolve candidates via optimization. Each job samples thousands of sequences, filters to roughly ten lab-ready designs, and runs at scale on Google Kubernetes Engine using spot TPU v6e, yielding a reported 3.65x performance-per-dollar advantage over H100 GPUs.

🛡️ CISA released an alert about a widespread software supply chain compromise affecting the npm registry: a self-replicating worm called 'Shai-Hulud' has compromised over 500 packages. The actor harvested GitHub Personal Access Tokens and cloud API keys for AWS, Google Cloud, and Azure, exfiltrating them to a public repository and using them to publish malicious package updates. CISA recommends immediate dependency reviews, credential rotation, enforcing phishing-resistant MFA, pinning package versions to releases before Sept. 16, 2025, hardening GitHub settings, and monitoring for anomalous outbound connections.

🔒 Google has added integrated security dashboards to GCE and GKE consoles, powered by Security Command Center. The dashboards surface top security findings, vulnerability trends, CVE prioritization, and container/workload misconfigurations informed by Google Threat Intelligence and Mandiant analysis. Teams can remediate misconfigurations, prioritize patches, and monitor threats directly in their compute and cluster consoles. Full vulnerability and threat widgets require upgrading to SCC Premium (30‑day trial available).

🚀 Google Cloud announced a new suite of AI training courses for intermediate and advanced learners across technical and non-technical roles. The curriculum covers designing and managing AI infrastructure using GCE and GKE, fine-tuning models like Gemini, serverless inference with Cloud Run, and securing generative AI deployments. Hands-on labs teach building AI agents that securely connect to enterprise databases and rapid prototyping in Google AI Studio. Courses are available on Google Cloud Skills Boost to help learners future-proof their AI skills.

🤖 Seattle Children’s partnered with Google Cloud to build Pathway Assistant, a multimodal AI chatbot that turns thousands of pediatric clinical pathway PDFs into conversational, searchable guidance. Using Vertex AI and Gemini, the assistant extracts JSON metadata, parses diagrams and flowcharts, and returns cited answers in seconds. The tool logs clinician feedback to BigQuery and stores source documents in Cloud Storage, enabling continuous improvement of documentation and metadata.

🔍 This post introduces the third installment in Google Cloud’s Network Performance Decoded series, summarizing three whitepapers that examine TCP retransmission tuning, the effects of headers and MTU on effective throughput, and techniques to measure packets-per-second with netperf. The guidance highlights practical kernel tuning (for example, rto_min and thin linear timeouts), how protocol and cloud-specific headers reduce payload efficiency, and rigorous netperf methodologies for sizing tests and correcting skew when measuring PPS. While examples reference Google Cloud features such as Protective ReRoute, the recommendations are broadly applicable to cloud deployments seeking improved responsiveness and accurate benchmarking.

🤖 Mr. Cooper partnered with Google Cloud to develop CIERA, a modular agentic AI framework that assembles specialized agents to support mortgage servicing representatives and customers. The design assigns distinct roles — orchestration, task execution, data retrieval, memory, and evaluation — while keeping humans in the loop for verification and personalization. Built on Vertex AI, CIERA aims to reduce research time, lower average handling time, and preserve trust and compliance in regulated workflows.

🔒 Google Cloud presents its certified MSSP ecosystem as a way to modernize security operations by combining partner expertise with Google Cloud Security products. Partners accelerate deployments and migrations, shorten time to value, and augment limited internal teams with specialized talent and AI-enabled tooling such as Google Security Operations and Mandiant. By providing scalable, 24/7 managed detection and response, MSSPs can reduce manual alerts, lower operational costs, and protect workloads across on-premises and multicloud environments.

🚀 Google Cloud highlights how its differentiated AI tech stack is accelerating startup innovation worldwide, with nine of the top ten AI labs, most AI unicorns, and more than 60% of generative AI startups using its platform. Startups are leveraging Vertex AI, TPUs, multimodal models like Veo 3 and Gemini, plus services such as AI Studio and GKE to build agents, generative media, medical tools, and developer platforms. Programs like the Google for Startups Cloud Program provide credits, mentorship, and engineering support to help founders scale.

🔒 A centralized proxy architecture on Google Cloud can secure remote Model Context Protocol (MCP) servers by intercepting tool calls and enforcing consistent policies across deployments. Author Lanre Ogunmola outlines five core MCP risks — unauthorized tool exposure, session hijacking, tool shadowing, token/theft and authentication bypass — and recommends an MCP proxy (Cloud Run, GKE, or Apigee) integrated with Cloud Armor, Secret Manager, and identity services for access control, secret scanning, and monitoring. The post emphasizes layered defenses including Model Armor for prompt/response screening and centralized logging to reduce blind spots and operational overhead.

🧰 MCP Toolbox now includes comprehensive Firestore tools that let AI assistants connect directly to Firestore from environments like Gemini CLI and other MCP-compatible interfaces. Built on the Model Context Protocol, these pre-built tools support document reads, collection queries, targeted updates, and security-rules validation to accelerate debugging, testing, and maintenance for NoSQL applications. Developers can perform complex queries and targeted updates in natural language, validate security rules before deployment, and reduce context switching between consoles and emulators. The release is accompanied by docs, quick start guides, a GitHub repo, and community channels to help teams adopt the features quickly.

📡 Over the past decade, Google Cloud evolved GKE pod networking from basic kubenet and route-based clusters to VPC-native alias IPs and the eBPF-powered Cilium Dataplane V2, improving performance, scalability, and observability. The platform now supports extreme-scale AI workloads with multi-NIC, terabit throughput, and persistent IPs for stateful functions. Looking forward, Google is exploring the Kubernetes Network Driver and the DRANET reference to expose node-level network resources via Dynamic Resource Allocation.

🚀 Google Cloud announced BigQuery performance and usability enhancements to accelerate generative AI inference. Improvements include >100x throughput for first-party text generation and >30x for embeddings, plus support for Vertex AI Provisioned Throughput and dynamic token batching to pack many rows per request. New reliability features—partial-failure mode, adaptive traffic control, and robust retries—prevent individual row failures from failing whole queries and simplify large-scale LLM workflows.

🚀 California is partnering with Google Cloud to modernize state and local services by applying AI, security, and infrastructure solutions. Agencies such as Covered California use Document AI, Assured Workloads, and AI-driven security to speed eligibility decisions and protect sensitive data. Universities including UCR and Caltech are using Vertex AI and AI-optimized HPC for research acceleration. Workspace, Gemini, and Agentspace are cited as productivity and information-management enablers.

🔍 Google Cloud and consultancy 28Stone published benchmarks showing the C3 machine series delivers low-latency, low-jitter performance suited to electronic trading. Tests using DPDK and replayed CME Group equity pcaps reported decision latencies as low as 1.5 µs (P50) and 3.5 µs (P99) and demonstrated consistent profiles at up to 100× data rates. The results highlight sub‑50 µs end-to-end round-trip P99 performance, high throughput with up to 200 Gbps per VM networking, and network innovations such as Titanium offload and Cloud WAN layer‑2 connectivity.

🪱 Researchers have identified a self-propagating npm worm dubbed Shai-Hulud that injects a 3MB+ JavaScript bundle into packages published from compromised developer accounts. A postinstall action executes the bundle to harvest npm, GitHub, AWS and GCP tokens and to run TruffleHog for broader secret discovery. The worm creates public GitHub repositories to dump secrets, pushes malicious Actions to exfiltrate tokens, and has exposed at least 700 repositories; vendors urge rotation of affected tokens.

🛡️ Researchers disclosed a campaign that trojanized more than 40 npm packages, including the popular tinycolor, embedding self-replicating credential-stealing code. The malware harvested AWS, GCP and Azure credentials, used TruffleHog for secrets discovery, and established persistence via GitHub Actions backdoors. Affected packages were removed, but developers are urged to remove compromised versions, rebuild from clean caches, and rotate any exposed credentials.

🛡️ DNS Armor is a cloud-native DNS security service from Google Cloud, built in partnership with Infoblox, that provides preemptive detection and mitigation of DNS-based threats for Google Cloud workloads. By intercepting internet-bound DNS queries and inspecting them in real time with Infoblox Threat Defense, it identifies malicious and high-risk domains, C2 activity, DNS tunneling, DGA patterns and evasive techniques such as fast-flux, and forwards detailed logs to Cloud Logging, Security Command Center, or SIEMs. Delivered as a turnkey managed service with no VMs and no impact to Cloud DNS, DNS Armor is enabled at the project level for granular protection and is available now in preview.

🧭 Google Cloud has expanded the Data Science Agent in Colab Enterprise notebooks to support BigQuery ML, BigQuery DataFrames and Spark, enabling large-scale data transformation, model training, and inference directly on BigQuery or via Serverless for Apache Spark. The agent can now auto-retrieve BigQuery table metadata and lets you add tables via an @ mention from your current project to provide prompt context. To invoke frameworks, include keywords such as BigQuery ML, BigFrames, or PySpark; sample prompts are provided to guide forecasting, supervised learning, and dimensionality reduction workflows. Notable limitations: generated PySpark targets Spark 4.0 and @ mentions only search the current project; BigQuery improvements are available now in BigQuery notebooks and coming soon to Vertex AI.

🚀 Google expanded BigQuery ML to generate embeddings from Gemini and over 13,000 open-source text-embedding models via Hugging Face, all callable with simple SQL. The post summarizes model tiers to help teams trade off quality, cost, and scalability, and introduces Gemini's Tokens Per Minute (TPM) quota for throughput control. It shows a practical workflow to deploy OSS models to Vertex AI endpoints, run ML.GENERATE_EMBEDDING for batch jobs, and undeploy to minimize idle costs, plus a Colab tutorial and cost/scale guidance.