ShShell.com - Latest AI News, Agentic AI, AI Tools, Prompts and Courses

Latest wire

NVIDIA's City AI Push Turns Public Services Into an Agent Reliability Test NVIDIA Puts Agentic AI on the Search for Extraterrestrial Signals C-BRAIN Uses Federated AI to Search Dementia Research Without Moving Every Dataset NVIDIA Cosmos Makes Physical AI Depend on the Data Factory Behind the World Model NVIDIA Minerva Treats AI Factory Reliability as a Yield Problem, Not an Alerting Problem NVIDIA Omniverse Makes Frontier AI Agents Useful Before They Touch the Real World

Filtered edition

Infrastructure & Robotics

Clear filter

Infrastructure

NVIDIA's City AI Push Turns Public Services Into an Agent Reliability Test

NVIDIA says cities are scaling physical and agentic AI services; the harder question is how public agencies verify, govern, and recover from automated decisions.

18h ago13 min readBy Sudeep DevkotaAI News

Infrastructure

NVIDIA Puts Agentic AI on the Search for Extraterrestrial Signals

NVIDIA’s account of AI agents helping search for extraterrestrial signals shows how scientific discovery changes when software can filter, rank, and revisit evidence.

Oct 3117 min readBy Sudeep DevkotaArtificial Intelligence

Infrastructure

C-BRAIN Uses Federated AI to Search Dementia Research Without Moving Every Dataset

The C-BRAIN consortium is combining governed institutional data, NVIDIA systems, and open AI tools to shorten the path from dementia research evidence to testable hypotheses.

Oct 3015 min readBy Sudeep DevkotaAI & LLMs

Infrastructure

NVIDIA Cosmos Makes Physical AI Depend on the Data Factory Behind the World Model

Oct 3015 min readBy Sudeep DevkotaRobotics

Infrastructure

NVIDIA Minerva Treats AI Factory Reliability as a Yield Problem, Not an Alerting Problem

Oct 3015 min readBy Sudeep DevkotaAI Infrastructure

Chronological feed

Latest in Infrastructure & Robotics

The newest reporting and analysis, ordered by publication time and labeled by editorial desk.

View all

Infrastructure

NVIDIA Omniverse Makes Frontier AI Agents Useful Before They Touch the Real World

NVIDIA is pairing frontier models with Omniverse physics, rendering, and OpenUSD so engineers can test buildings, turbines, and physical-AI workflows in simulation.

Oct 3016 min readBy Sudeep DevkotaAI Engineering

Infrastructure

Google and Microsoft Pause Their AI Capex Forecasts, Forcing the Build-Out to Answer for Itself

Google and Microsoft are holding back fresh AI infrastructure forecasts as investors demand evidence that massive compute spending produces durable returns.

Oct 2916 min readBy Sudeep DevkotaArtificial Intelligence

Infrastructure

Retell’s 100 Million Monthly Calls Put Voice Agents on a Reliability Clock

Retell’s OpenAI-powered voice-agent scale shows why latency, interruption handling, and escalation matter more than a convincing demo.

Oct 2821 min readBy Sudeep DevkotaAI Business

Infrastructure

Schneider Electric’s Modular AI Factory Design Targets the Time Lost Between Chips and Power

Schneider Electric’s new modular reference design for NVIDIA Vera Rubin infrastructure treats AI data-center deployment time as an engineering bottleneck.

Oct 2313 min readBy Sudeep DevkotaAI News

Infrastructure

Google and Voltus Put a One-Gigawatt Virtual Power Plant Behind the AI Boom

Google says it is working with Voltus on a one-gigawatt virtual power plant, linking AI-era electricity demand to flexible grid operations.

Oct 2222 min readBy Sudeep DevkotaAI Infrastructure

Infrastructure

Google’s Coral NPU Standard Moves Edge AI Toward Portable Silicon

Google’s Coral NPU initiative aims to make edge AI portable across wearable and industrial chips by separating model software from one accelerator design.

Oct 2114 min readBy Sudeep DevkotaArtificial Intelligence

Infrastructure

llm-d 1.0 Treats AI Inference as a Scheduling Problem, Not Just a GPU Problem

llm-d 1.0 brings open scheduling, routing, and cache-aware controls to large-model serving, making inference economics an infrastructure discipline.

Oct 2114 min readBy Sudeep DevkotaArtificial Intelligence

Infrastructure

NVIDIA Opening Run:ai Turns AI Factory Scheduling Into Shared Infrastructure

NVIDIA’s decision to open source Run:ai as part of its DSX AI Factory platform targets a problem that is easy to miss in model headlines: expensive acceler

Oct 2013 min readBy Sudeep DevkotaArtificial Intelligence

Infrastructure

NVIDIA’s UK Startup Investments Show the Next AI Race Is About Ecosystem Density

NVIDIA’s new investments in the United Kingdom startup ecosystem are a signal about where infrastructure companies expect durable advantage to form: not on

Oct 2014 min readBy Sudeep DevkotaArtificial Intelligence

Infrastructure

NVIDIA’s Nemotron 3.5 Super VL and Bolt Put Agent Efficiency Ahead of Model Size

NVIDIA’s new open Nemotron 3.5 Super VL and Bolt models shift the agent race toward efficient multimodal inference, reproducible recipes, and deployment cost.

Oct 1914 min readBy Sudeep DevkotaAI Engineering

Infrastructure

NVIDIA and Cisco Put Pluggable Optics at the Center of AI Factory Scale

NVIDIA and Cisco are introducing a pluggable optics design for AI factories, shifting the bottleneck from GPUs toward network power, reach, and serviceability.

Oct 1813 min readBy Sudeep DevkotaAI Infrastructure

Infrastructure

NVIDIA’s Nemotron Deep-Agent Result Makes Evaluation Harnesses the Real AI Product

NVIDIA’s Nemotron result with LangChain’s Deep Agents harness shifts attention from model scores to the workflow that produces reliable agent outcomes.

Oct 1722 min readBy Sudeep DevkotaAI Agents and LLMs

Infrastructure

Liquid AI Releases LFM2.5-24B-A2B, a 24B-Parameter Model Built for On-Device Agents

Liquid AI's LFM2.5-24B-A2B release makes local agent design a memory, latency, and control problem rather than a cloud-model shopping exercise.

Oct 1414 min readBy Sudeep DevkotaAI Engineering

Infrastructure

Microsoft’s Maia 300 Makes AI Inference a Power and Supply-Chain Contest

Microsoft’s Maia 300 inference accelerator shows why serving AI at scale now depends on watts, rack design, software compatibility, and supply-chain control as much as raw model speed.

Oct 1414 min readBy Sudeep DevkotaAI Infrastructure

Infrastructure

NEURA’s Neuron Foundation Model Makes Physical AI a Deployment Bet

NEURA Robotics' Neuron model highlights the difficult middle layer of physical AI: turning multimodal perception and language into safe, repeatable actions on real machines.

Oct 1414 min readBy Sudeep DevkotaRobotics

Infrastructure

CoreWeave’s 240 MW India Expansion Makes AI Infrastructure a Locality Problem

CoreWeave’s reported AdaniConneX deployment in India shows how AI cloud growth now depends on power, land, connectivity, and regional demand as much as on accelerators.

Oct 1113 min readBy Sudeep DevkotaAI Infrastructure

Infrastructure

NVIDIA’s Reflection AI Talks Put Open Models on a Consolidation Clock

Reports that NVIDIA is considering a deeper investment or acquisition of Reflection AI reveal how open-weight model competition is becoming an infrastructure and distribution contest.

Oct 1114 min readBy Sudeep DevkotaAI Industry

Infrastructure

Robot Hands Are Becoming the Real Test of Embodied AI

Boston Dynamics’ new robot-hand work shifts the embodied-AI debate from humanoid spectacle toward dexterity, data, and reliable work.

Oct 1014 min readBy Sudeep DevkotaRobotics

Browse the full news archive