Models & Research
Nemotron Diffusion tests the one-token-at-a-time bottleneck
NVIDIA released tri-mode diffusion LLMs that switch between AR, diffusion, and self-speculation generation in one checkpoint.
Industry & Policy
SKILL.md is not documentation, it is the new signed supply chain
NVIDIA Verified Agent Skills treats agent skills as scanned, carded, and signed artifacts, pointing to a new supply-chain checkpoint for AI agents.
AI Agents
88-Core Vera Shows Agent Bottlenecks Moving To The CPU
NVIDIA Vera CPU deliveries show that AI agent competition is shifting beyond GPU inference toward CPU-bound sandboxes, tool calls, and RL evaluation.
AI Agents
AI-Q Skill and the Data Boundary for Research Agents
NVIDIA AI-Q agent skill lets Claude Code, Codex, and other harnesses delegate enterprise research to a local AI-Q server.
Models & Research
SANA-WM 2.6B asks what a one-minute world model really costs
NVIDIA SANA-WM claims 720p, 60-second world modeling from a 2.6B backbone. The real story is not video polish but the cost structure of open models.
AI Agents
Hermes Agent turns the local PC into a learning agent runtime
NVIDIA is positioning Nous Research Hermes Agent on RTX and DGX Spark as a local, always-on self-improving agent runtime.
Models & Research
NVIDIA Is Targeting the RL Training Loop
NVIDIA and Ineffable Intelligence are pointing the model race toward RL infrastructure for agents that learn from experience.