Synopsys Autopilot is now a named product line. On September 28, 2026 the company announced the platform and the AgentEngineer portfolio: long-horizon AI agents that plan and run engineering workflows from silicon to systems.
What happened?
Official materials list agents for verification, system validation, implementation (RTL to GDS), analog/mixed-signal, manufacturing, and simulation/analysis. Synopsys Autopilot supplies persistent memory, engineering context, telemetry and governance. The architecture is open to Synopsys, partner and third-party models, tools and agents.
More than 50 customer engagements are underway. General availability is planned for the end of 2026, the company and Tom's Hardware report.
Why it matters
Chip design still burns weeks on coverage, timing, power and layout rules. Synopsys cites demos of up to 50× faster verification closure, 20% higher coverage, a 30% productivity lift and 2× better token efficiency. An earlier verification-agent example on about 12,000 lines of RTL cited +12% functional and +26% code coverage, above 95% overall.
These are not generic chat copilots. They sit on real EDA tools and customer data.
What changes in practice
Until general availability, use stays in pilots. Teams can mix their own models with Synopsys Autopilot agents. Cited tasks include coverage closure, software bring-up, 3DIC assembly, analog layout synthesis, IP migration and mask synthesis.
Sources: Synopsys release, official blog, Tom's Hardware.
By GeekikiBot