The development cycle is shortened by 40%! Synopsys joins forces with AMD and Microsoft: using AI to accelerate chip design

Kuai Technology reported on August 13 that Synopsys announced the launch of a new autonomous intelligent AI workflow for the chip design field. This workflow was developed jointly with AMD and is now available for evaluation on the Microsoft Discovery platform. Artificial intelligence is fundamentally reshaping the engineering landscape. Synopsys is committed to building a comprehensive and open intelligent AI technology stack, aiming to improve the level of engineering autonomy and accelerate full-link product development from chips to systems. At the 2026 DAC Chips to Systems Conference held recently, Synopsys demonstrated a new fully autonomous workflow driven by AgentEngineer technology. This workflow marks that the application of agent AI has expanded from task automation to long-cycle project execution. The two Synopsys autonomous workflows released this time based on the Microsoft Discovery platform include: Fully autonomous debug closure workflow (Debug Closure Workflow). This workflow is driven by AI and can realize fully autonomous verification and root cause analysis (RCA). It relies on the Microsoft Discovery platform for orchestration and integrates domain-specific agents and task-level agents. It can automatically identify design defects, perform debugging tasks and accelerate the verification process, helping engineering teams quickly locate and solve problems, thereby improving chip quality. Early evaluation shows that this workflow can shorten the debugging cycle by 25%–40%, save weeks of engineering investment, and significantly improve overall efficiency. Completely autonomous implementation and closure workflow (Implementation and Closure Workflow) This autonomous workflow is based on Synopsys implementation agents and Fusion Compiler running on the Azure platform, integrating domain agents and task-level agents to achieve design implementation quality (QoR) optimization and automation of the convergence process. Preliminary results show that this workflow can further improve QoR performance. AMD is currently