As AI spreads across physical products such as cars, medical devices and industrial equipment, semiconductor design software company Synopsys presented a new product development paradigm. Synopsys said on June 29 it unveiled its Silicon-to-Systems vision after acquiring engineering simulation company Ansys.

Synopsys said the core of the approach is linking decision-making in the semiconductor design stage to system-level outcomes. It said cars are changing into products where competitiveness is determined by the computing platform, medical devices are evolving into compact data processors, and industrial equipment is shifting toward predicting failures on its own.

Synopsys said the traditional sequential development method of design, manufacturing, testing and revision has hit its limits. Errors found late in development sharply raise the cost of fixes, and problems discovered after semiconductor tape-out hurt both schedule and cost. It said shift-left methods, which verify designs at the system level before a product physically exists, are emerging.

AI is being used as a productivity tool that supports design space exploration, trade-off optimization and automation of repetitive tasks, the company said. It said as reinforcement learning, generative AI and agent-based workflows are gradually applied to design processes, engineers can focus on creativity and product differentiation rather than repetitive work.

Synopsys said digital twins are evolving beyond simple virtual models into "living systems" that reflect products in actual operation. It said data collected during operation is fed back into the design stage to continuously improve product reliability and performance and support the development of next-generation products.

A Synopsys official said, "Companies that succeed in the future will be those that can design these layers at the same time and resolve various trade-offs at an early stage."