Latest Tech News
Daily updates from AI, software, hardware and more.
NVIDIA pitches Vera Rubin for cheaper continuous AI post-training
NVIDIA says Vera Rubin is designed to lower the cost of continuous post-training for agentic AI systems.
GeForce NOW exits beta in India and adds Onimusha launch-day streaming
NVIDIA says GeForce NOW is now publicly available in India and will stream Onimusha: Way of the Sword at launch.
NVIDIA and Japan-backed Noetra plan 140MW AI factory for physical AI models
NVIDIA says a Japan-backed Noetra AI factory will use Vera Rubin systems to train physical AI and robotics models.
NVIDIA CUDA 13.3 adds GPU instruction aimed at faster cryptography workloads
NVIDIA says CUDA 13.3 adds clmad, a GPU instruction that speeds GHASH and some zero-knowledge proof arithmetic workloads...
NVIDIA adds smaller Jetson Thor modules for robotics and edge AI systems
NVIDIA introduced Jetson T3000 and T2000 modules for robotics and edge AI, with Thor architecture and Q1 2027 availabili...
NVIDIA opens AI decoder tools claiming 300x-plus quantum color-code error gains
NVIDIA released open AI decoder tools for quantum color codes, claiming major error-rate and runtime gains in a new benc...
NVIDIA outlines hardware-friendly LLM design rules for faster Blackwell inference
NVIDIA says LLM designers should align model dimensions, quantization, and parallelism with GPU hardware for faster infe...
NVIDIA says tuned Nemotron 3 Ultra stack cuts agent inference cost against closed models
NVIDIA says a tuned LangChain Deep Agents stack for Nemotron 3 Ultra improves open agent performance and lowers inferenc...
NVIDIA details Vera CPU push for agentic AI data centers
NVIDIA says its Vera CPU targets agentic AI bottlenecks around tool use, sandboxing, analytics, and GPU utilization.
NVIDIA and Hugging Face add Isaac robotics tools to LeRobot
NVIDIA and Hugging Face are adding Isaac GR00T 1.7 and Teleop workflows to LeRobot for open robotics development.
NVIDIA says open Nemotron models underpin 145 ICML 2026 papers
NVIDIA says 145 ICML 2026 papers cite Nemotron open models as open AI stacks spread into robotics, biology and agents.
NVIDIA says Blackwell confidential AI inference keeps up to 98% of baseline performance
NVIDIA says Blackwell confidential computing can protect AI inference with benchmark overhead generally below 8.1%.