AI Amplifying Chip Development Processes
The future of Electronic Design Automation (EDA) is being shaped by a significant shift towards AI-driven automation and multi-agent AI systems. This transformation is set to redefine chip design workflows, reshape the engineering workforce, and revolutionise the semiconductor industry.
Impact on Chip Design Workflows
AI is being increasingly integrated across the entire chip design cycle, from synthesis and floorplanning to formal verification, debugging, layout, manufacturing, and testing. This integration enables several advancements:
- Autonomous design space exploration that optimises power, performance, and area (PPA) more efficiently than manual approaches.
- AI-assisted verification targeted towards corner cases and prioritised test generation, accelerating closure and improving coverage for complex SoCs.
- Natural-language-driven design tools, such as ChipGPT, that translate specifications into hardware description languages (HDL) like Verilog, speeding up the transition from conceptual to functional design.
- AI-enabled layout and routing tools reducing error rates and development time.
- Predictive analytics in manufacturing and testing for yield enhancement and defect detection.
Multi-agent AI systems, highlighted at DAC 2025, represent the next evolution in EDA. These agents collaboratively operate within EDA toolchains to catch low-level errors, accelerate information retrieval, and unify workflows across multiple tools and data sources. Siemens EDA’s AI system exemplifies this with open, integrative platforms supporting both standalone AI tools and end-to-end design flows.
Impact on the Engineering Workforce
AI is expected to reduce repetitive, manual tasks and information hunting, currently consuming up to 40% of an early-career engineer's time. This allows engineers to focus on higher-value, creative, and innovative activities.
Multi-agent AI systems are envisioned as "employees of the future," complementing but not fully replacing human judgement and domain expertise. Human oversight remains critical for decision-making and handling novel design challenges.
AI augmentation fosters more efficient collaboration and faster iteration cycles, enabling engineers and architects to push chip innovation with less friction and delay.
Impact on the Industry as a Whole
The convergence of AI-driven automation, cloud-native EDA tools, chiplet-based architectures, and security-aware design is reshaping ASIC development at a fundamental level.
AI accelerates time-to-market and enhances design quality, helping semiconductor companies to cope with the rising complexity of SoCs and market demands.
Increased adoption of AI frameworks in EDA promotes competitive advantage, drives innovation, and creates new roles oriented towards human-AI collaboration in chip development.
Privacy and data confidentiality concerns guide AI deployment strategies; on-premise AI solutions are preferred to protect proprietary designs and data, ensuring trust and security.
In summary, AI’s role in EDA is evolving from assistive tools to fully integrated, collaborative AI ecosystems, drastically improving chip design efficiency, allowing engineers to focus on strategic innovation, and driving industry-wide transformation towards faster, smarter, and more secure semiconductor design and manufacturing processes.
The transition to AgentEngineer technology, as outlined by Synopsys President and CEO Sassine Ghazi, will allow design teams to re-engineer engineering and take advantage of the latest AI tools and innovations.
AI assistants free engineers from mundane aspects of chip design, allowing them to focus on architectural innovations, novel uses of EDA tools, and higher-level problem solving.
Agentic AI is the next wave of EDA automation and optimization, offering truly automated decision-making and orchestration with minimal human intervention.
Current progress and outlook are overwhelmingly positive, with the integration and use of AI tools already delivering noticeable results, and agentic AI on the horizon, marking the semiconductor industry's entry into a new era of automation and innovation.
Level 5 of automation, or the "autopilot," will offer high-level decision-making, fully autonomous reasoning, and complex planning capabilities, allowing human engineers to simply enter the product specification and an entire subsystem to be created automatically.
AI is being used in tasks such as analog design, test and verification, RTL and verification collateral generation, timing analysis, and design rule checking.
- The integration of AI across the chip design cycle, including tasks like autonomous design space exploration, AI-assisted verification, natural-language-driven design tools, and AI-enabled layout and routing tools, is redefining workflows in chip design.
- The advent of multi-agent AI systems, collaboratively operating within EDA toolchains, represents the future of the semiconductor industry, as they are expected to catch low-level errors, accelerate information retrieval, and unify workflows, complementing human judgement and expertise.
