Nvidia and ChipAgents are expanding their work together on AI agents that help design and verify computer chips. The two companies announced the broader collaboration on July 26, 2026.
What Nvidia and ChipAgents are building
At the center of the deal is Renoir, the model and agent system built by ChipAgents. The partnership is meant to scale Renoir using Nvidia’s AI software stack and computing infrastructure.
Renoir has been fine-tuned on semiconductor datasets. According to the research findings, Renoir has reportedly reached performance close to leading frontier models. The headline cost claim is aggressive: operating costs potentially as low as 22% of Claude Fable 5.
That matters because chip design is not a one-prompt job. Agents run long chains of tasks, and every step costs compute. Cheaper inference makes it more realistic to let agents grind through those workflows at scale.
Renoir is aimed at end-to-end chip workflows. That covers RTL generation, verification plans, assertions, simulations, and IP protection.
Nvidia executive Da Yang described AI agents as:
“a new class of digital workers”
Yang positioned these agents as essential for handling the growing complexity of modern engineering workflows.
The ChipAgents track record so far
ChipAgents was founded in 2024 by William Wang, a professor at UCSB. The company builds autonomous agents that can manage multiple stages of semiconductor design, from writing the specification to performing root-cause analysis when something breaks.
In January 2025, ChipAgents reported a 97.4% pass rate on Nvidia’s VerilogEval-v2 benchmark. It completed 152 out of 156 tasks.
ChipAgents’ technology is deployed at more than 120 semiconductor companies, including MediaTek and Micron. The company claims its tools cut design and verification cycles by more than 50%.
ChipAgents has raised approximately $131-134 million across funding rounds, with Series A activity running between October 2025 and July 2026. Backers include Bessemer and MediaTek.
MediaTek’s dual role stands out. It is both a customer and an investor.
Why the chip industry wants AI agents
The partnership lands in the middle of a broader 2026 trend. At industry events such as DAC, the Design Automation Conference, integrating agentic AI into electronic design automation tools has been a major theme.
Two pressures are driving that push. Chips keep getting more complex, and the industry faces a shortage of skilled engineers to design them. Verification already eats more than 70% of design project time.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.