Ex-NVIDIA AI lead's chip startup Agentrys raises $19.1M seed
What's the deal? Agentrys, a startup automating chip design, has raised $19.1 million in an oversubscribed seed round led by Etna LabsDealroom has a profile for this one. Try Dealroom →. The total funding reaches $24.5 million, including a $5.4 million pre-seed led by MediaTekDealroom has a profile for this one. Try Dealroom →, its first strategic backer.
Who's behind it? Founder and chief executive officer Mark Ren spent nearly three decades in Electronic Design Automation (EDA) and AI research at NVIDIA Research and IBM ResearchDealroom has a profile for this one. Try Dealroom →. He led ChipNeMo, the first industrial large language model for chip design.
What's the endgame? Agentrys is defining a category it calls Agentic Design Automation (ADA), which shifts chip work from single-task EDA tools to systems that learn from and improve entire engineering workflows. The new funding will go toward recruiting, agent-native tooling, and expanding customer work across verification and physical design.
Why it matters: Chip design still leans on scarce experts and thousands of manual engineering hours per chip. Agentrys targets the workflows that decide whether chips in the $800 billion-plus semiconductor industry ship on time and hit performance targets.
Agentrys pitches itself as an open platform customers build on, rather than a fixed set of vendor-controlled agents. Its design intelligence layer learns continuously from customer data, usage, and evaluation signals. "That's the difference between renting capability everyone else can rent and building an asset that compounds for you alone," Ren said.
What's the investor case? Etna Labs, a research-driven firm that has backed more than 20 AI companies, sees chip design as fertile ground for recursive self-improvement because results "can be objectively evaluated with rigorous engineering tools and metrics," members of its investment team said.
The signal: The round sits in the 95th percentile of seed rounds in its geography over the trailing 48 months, a sign of investor appetite for applying AI agents to high-value, verifiable technical domains — with chip design an early proving ground.
Read more: morningstar.com
Image credit: Agentrys
