Synopsys Introduces New AI Platform to Automate End-to-End Chip Design Workflows

Synopsys introduced its AgentEngineer portfolio and Autopilot Platform, aiming to automate engineering workflows across chip design and production, from verification and implementation to analog design, manufacturing, and simulation. The platform combines long-horizon agents that coordinate work with task-specific agents, while providing orchestration, memory, governance, and enterprise security controls; Synopsys says engineers retain approval checkpoints. The company reports more than 50 customer engagements and plans general availability by the end of 2026. Customers including Fujitsu, Intel, MediaTek, Samsung, and AheadComputing described work with the technology, with Fujitsu reporting a 10% to 30% productivity increase in RTL code generation.
Synopsys reported engagement results of up to 50 times faster verification closure, 20% higher coverage, and twice the token efficiency; these are company-reported results, not independently verified figures.
Specific task-level agent capabilities include autonomous coverage closure, software bring-up and validation, multi-die 3DIC assembly, power-performance-area closure, analog layout synthesis and design migration, and mask synthesis.
Synopsys described an open lane for customer-built agents, alongside the company’s portfolio, and the platform supports customer choice across infrastructure, models, data, and tools.
NVIDIA said it is combining its accelerated computing and Agent Toolkit with Synopsys’ platform to build agents focused on design quality and development acceleration; TSMC cited AI-based power-integrity analysis workflows for multi-die architectures.
Publishers
18
Articles
60
Reach
78