Solving the $100k License Tax: A Convergence of Industry Perspectives on Agentic AI for chip design

In our previous post, we discussed the gap between AI "demos" and production-grade silicon. Today, we address the primary friction point preventing the industry from scaling AI-driven verification: The License Tax.

The Math of Iteration

"Agentic AI"—where autonomous loops generate, test, and fix RTL—relies on high-frequency iteration. For an AI agent to "learn" a fix or optimize a module, it may need to run hundreds or thousands of cycles through a verification engine.

In a traditional environment, this is an economic impossibility. If every iterative trial pings a commercial tool seat—often costing upwards of $100,000 per license—the cost of the "agentic workforce" scales faster than its productivity. To make AI ROI viable, we must decouple the high-volume iterative cleaning from the high-cost final sign-off.

The Production Milestone: Verilator + UVM is becoming a reality. AsFigo and VerifWorks have already ported 50+ production-grade UVCs to Verilator. While these currently support the first few tests with documented limitations and workarounds, they represent a functional "Utility" layer that was previously unavailable for high-frequency AI iteration.

Building the Guardrails

The solution lies in a hybrid flow. At AsFigo Technologies, we’ve focused on building the "Ground Truth" necessary for these AI agents to operate safely and affordably. By utilizing open-source engines like Verilator and Slang, we provide a low-overhead sandbox for AI agents to iterate.

Crucially, this iteration is governed by specific guardrails such as SVALint and UVMLint. These tools, built on top of Google Verible and Slang, act as the gatekeepers—ensuring that AI-generated code meets rigid industrial standards before it is ever promoted to a commercial simulator.

A Convergence of Industry Perspectives

This isn’t a theoretical shift; it’s a practical necessity being addressed by a diverse group of industry veterans and innovators. We are hosting a private co-located meeting during the week of DVCon U.S. 2026 to explore how this stack is being viewed and deployed across the following perspectives:

  • Deepak Tala (SmartDV): The commercial reality of maintaining high-revenue IP quality in an evolving tool landscape.
  • Shashank Chaurasia (MooresLabAI): Real-world implementation of "tool-in-the-loop" agents within a VC-backed startup pipeline.
  • Akash Levy (Silimate): Exploring "AI-Native" flows for real-time PPA and functional feedback in VC-backed innovation.
  • Clifford Cummings (Sunburst Design): A veteran "Guru" perspective on the talent pipeline and expert education in an open-source world.
  • Yatin Trivedi (Capgemini): The enterprise view on managing mass AI deployment and risk for global design service teams.
  • Asif ETV (HPC Infra): The architectural requirements for secure, professional-grade infrastructure in the enterprise AI era.

Event Logistics

This is an independent, private event hosted by AsFigo Technologies and co-located with the conference. To ensure a focused technical environment, registration is mandatory, though the session is free for attendees.

Location: Ballroom C, Hyatt Regency, Santa Clara, CA
When: Monday, March 2nd, 2026 | 6:00 PM – 7:00 PM PST
Registration: [Link to be added]

Comments

Popular posts from this blog

Join Us in Cambridge - Advancing Chip Verification with UVMLint & SVALint

SVALint Technical Meetup – Reading, UK

Call for Collaboration: Seeking Public UVM Environments for Verilator 5.0+ Porting