Moving Beyond AI-Assisted Design – Agentrys Shows the Way with Agentic Loop Engineering

<p style="font-weight: 400;">At Hot Chips 2026, OpenAI presented Jalapeño, an inference ASIC built with AI in the design loop. As reported from the session by <a href="https://www.servethehome.com/openai-jalapeno-asic-at-hot-chips-2026/">ServeTheHome</a> and <a href="https://www.tomshardware.com/tech-industry/artificial-intelligence/hot-chips-2026-openais-jalapeno-ai-asic-unpacked-accelerator-developed-using-ai-achieves-efficiency-and-throughput-gains-against-power-hungry-blackwell">Tom’s Hardware</a>, OpenAI showed measured block-level wins over a human baseline — 56% on a BF16 multiply, alongside a 10% smaller matrix-unit area — within a roughly nine-month path from&#8230; <a href="https://semiwiki.com/eda/agentrys/375112-agentrys-shows-the-way-with-agentic-loop-engineering/" class="read-more">Read More </a></p> <p>The post <a href="https://semiwiki.com/eda/agentrys/375112-agentrys-shows-the-way-with-agentic-loop-engineering/">Moving Beyond AI-Assisted Design – Agentrys Shows the Way with Agentic Loop Engineering</a> appeared first on <a href="https://semiwiki.com">SemiWiki</a>.</p>