AI chip design

Nvidia has announced a watchdog chip designed to sit alongside every AI agent, monitoring its behaviour in real time and stepping in when the agent acts outside its intended limits. The move positions Nvidia to sell safety hardware as autonomous AI systems spread through industry, and it is drawing debate among developers about oversight, cost, and who controls the safeguards.

Reports describe how artificial intelligence is increasingly used to design computer chips, with AI assisting at every stage of semiconductor development, from electronic design automation software to automated layout and verification. OpenAI's internal chip effort, referred to as Jalapeño, is cited as an example of AI-driven silicon design, marking a shift toward chips largely designed by machines.

The Chips Joint Undertaking has launched funding worth €80 million to support the development of AI chips and compute systems in Europe. The call aims to strengthen European capabilities in semiconductor design and computing hardware for artificial intelligence, backing research and innovation projects across the bloc. The move is part of broader efforts to reduce reliance on foreign chip suppliers.

Chip startup EVAS has raised $295 million to expand its work on RISC-V-based AI architecture. The funding will support development of open-standard processor designs aimed at artificial intelligence workloads, a growing alternative to proprietary architectures. The announcement is drawing attention from the semiconductor and AI hardware communities tracking investment in open-source chip technology.

Synopsys has introduced AgentEngineer, an 'Autopilot' platform designed to develop chips autonomously using AI agents. The company says the platform will reach general availability by the end of 2026. The announcement is drawing attention in the semiconductor industry as chipmakers look to AI to speed up increasingly complex and costly chip design workflows.

NVIDIA has introduced an in-chip watchdog feature designed to monitor and constrain rogue AI agents running on its hardware, with roughly 100 companies reportedly joining the initiative. The move reflects growing industry concern that autonomous AI agents could act unpredictably or beyond their intended scope, prompting silicon-level safety and monitoring mechanisms.