Homechevron_rightArchiveschevron_rightTechnology

Archives

Explore Stories by Date

21 Oct 2026

Model Routing Slashes AI Inference Costs While Maintaining Quality

Organizations running AI models face mounting inference costs, but model routing offers a solution. By dynamically selecting the most appropriate AI model based on task complexity, companies achieve 20-95% cost reductions. Major vendors like Databricks, NVIDIA, and Snowflake now offer routing solutions, though teams must avoid the Pareto trap that can turn savings into customer retention losses.

2 Sources

GM Unveils AI-Powered Autonomous Driving Push With OnStar AI and Eyes-Off Highway Tech by 2028

General Motors announced sweeping AI and autonomous driving initiatives at its GM Forward event, including OnStar AI assistant launching in 2027 Cadillac Lyriqs and hands-free, eyes-off highway driving across its lineup by 2028. The automaker is deploying a 40-times more powerful central computing unit and cutting vehicle development time in half to compete with Chinese manufacturers.

15 Sources

Gartner Unveils 10 Critical AI Trends Reshaping Enterprise Technology Through 2027

Gartner has released its strategic technology trends for 2027, highlighting how autonomous AI agents and specialized models will transform enterprise operations. The research predicts that by 2030, over 70% of enterprise voice interactions will be handled autonomously, while 25% of security breaches will trace back to unmanaged AI agent privileges.

3 Sources

AI Models Achieve 90% Success Rate Planning Cyberattacks on Power Grids and Water Systems

Research from cybersecurity firm Frenos reveals AI models can successfully construct attack plans against critical infrastructure in roughly 90% of attempts. The findings expose vulnerabilities in power grids and water systems as AI-powered cyberattacks accelerate threat timelines and lower barriers for sophisticated infrastructure attackers.

3 Sources

NVIDIA and Nokia Transform Telecom Networks Into Distributed Intelligence Platforms With AI-RAN

NVIDIA and Nokia are partnering with global telecom operators including T-Mobile, Deutsche Telekom, Orange, Elisa, and BT to deploy AI-RAN technology. T-Mobile completed the world's first AI-RAN call on a commercial network in Bellevue, Washington. The initiative transforms telecom networks into distributed intelligence platforms by integrating NVIDIA accelerated computing into centralized data centers, metro sites, and cell sites.

5 Sources

NVIDIA DSX Enables European AI Factories to Slash Water Use and Cut Power Consumption by 30%

NVIDIA unveiled how European infrastructure leaders are adopting its DSX AI Factory Platform to build efficient AI factories with near-zero cooling water consumption. The platform can save 54 million liters daily at 2-gigawatt scale while using up to 30% less power for the same AI output.

3 Sources

Googlebooks Launch With Gemini AI Promises But Deliver Mixed Results for Android Laptop Users

Google unveiled its premium Android laptops called Googlebooks starting at $899, featuring Gemini AI integration and Android phone connectivity. However, early testing reveals significant limitations in AI functionality, app compatibility issues, and missing enterprise features that Chromebooks currently offer, raising questions about their readiness for widespread adoption.

3 Sources

NVIDIA Expands DGX Spark to 8-Way Clusters with 1TB Memory, Boosts AI Inference by 40%

NVIDIA unveiled major updates to its local AI hardware systems at GTC Berlin, expanding DGX Spark to support 8-way clustered configurations with up to 1TB of unified memory. The company also delivered a 40% boost in AI inference performance across DGX and RTX systems, alongside new model support and enterprise management integrations.

2 Sources

AMD and Intel merge CPU AI libraries to create unified PyTorch code path for x86 processors

AMD and Intel are merging their CPU optimization libraries for PyTorch into a single oneDNN framework. AMD's ZenDNN optimizations will integrate into Intel's oneDNN, creating one unified PyTorch code path for AI inference on CPUs. The move eliminates software fragmentation across x86 CPUs and simplifies AI development for developers working across both processor families.

3 Sources

OpenAI brings AI models to Google Cloud, completing expansion across major cloud providers

OpenAI announced its frontier AI models, including GPT-6 Astra, Sol, and Luna, will launch on Google Cloud in the coming weeks. The strategic partnership marks OpenAI's expansion to all three major U.S. cloud providers after Microsoft's exclusive license ended in April. Enterprise customers can now access OpenAI models through Google Cloud Marketplace alongside Gemini and Anthropic Claude.

15 Sources

Nvidia DSX POD Slashes AI Factory Deployment From 6 Months to Just 14 Days With Vera Rubin Racks

Nvidia unveiled DSX POD, a container-based AI factory that cuts datacenter deployment time from six months to 14 days. Each modular unit packs seven Vera Rubin racks with 504 GPUs, factory-tested and ready to connect. Partners including HPE, Dell, and Schneider Electric will deliver the plug and play systems starting early 2027.

10 Sources

NVIDIA Expands Isaac Platform With Agile ONE SN Humanoid, Isaac Lab 3.0 for General-Purpose Robotics

NVIDIA announced a major expansion of its Isaac robotics development platform at GTC Berlin, introducing the Agile ONE SN humanoid reference platform built on Isaac GR00T and Jetson Thor. The update includes Isaac Lab 3.0, Isaac OS runtime, and over 2,000 simulation-ready assets, cutting robot development setup time from weeks to days.

3 Sources

Nvidia Introduces OpenShell Architecture to Contain and Monitor AI Agents Across Enterprise Systems

Nvidia launched OpenShell Reference Architecture at GTC Berlin to address AI agent security concerns. The full-stack platform combines Vera Rubin infrastructure, Vera CPUs, and BlueField-4 DPUs running Sentry software to contain, monitor and enforce boundaries on autonomous agents within milliseconds.

13 Sources

Eternal Unveils Dark.ai, an End-to-End Encrypted AI Assistant for Sensitive Conversations

Zomato and Blinkit parent Eternal announced Dark.ai, a privacy-focused AI platform with end-to-end encryption that prevents even the company from reading user conversations. Founder Deepinder Goyal positions the AI assistant as a solution for power users including founders, researchers and enterprises who need to discuss sensitive information without concerns about data retention or access.

7 Sources

ElevenLabs Launches Voice AI Models With 98.8% Speech Reasoning Accuracy for Real-Time Conversations

ElevenLabs unveiled Eleven Realtime, two speech-to-speech AI models designed to simplify voice agent development. The v1 Max achieved 98.8% speech reasoning accuracy in independent benchmarks, while v1 Mini excels at conversational dynamics. Both models support over 90 languages and 17,000 voices, targeting developers building AI assistants and customer support systems.

2 Sources

Get all updates into your inbox 💌, Stay Up-to-date for Free 🥳