Cadence adds machine learning to electrical simulation

Cadence Design Systems has used machine-learning techniques originally developed for its Cerebrus tool to build software that can speed up the multiphysics analysis of interconnects and 3D layouts.

According to Cadence, the Optimality Explorer tool delivers designs optimized for signal integrity and electromagnetic compatibility on average ten times faster than traditional manual methods that rely on parametric sweeps and similar techniques.

Like Cerebrus, the system explorer applies machine learning to existing tools to support optimization flows, in this case the Clarity 3D Solver, used for 3D electromagnetic analysis, and Sigrity X software for signal-integrity and power-integrity.

Rather than rely on extensive parametric sweeps, which may involve nested loops involving ten or more variables, the explorer uses machine learning to guide the user to more productive simulation targets. Cadence says the time savings on more complex parametric optimizations can be a hundredfold.

The company claims the tool, based on design parameters provided by the designers, helps them quickly determine optimum electrical performance, and avoid suboptimal local minima and maxima, as well as mapping variations for additional consideration. The tool can also be used with the company’s multiphysics technologies.

“For years, optimization at the system level has been extremely inefficient based on a human-intensive workflow of design/prototype/test/refine and eventual manufacturing,” said Ben Gu, vice president of R&D for the multiphysics system-analysis business unit at Cadence. “With Optimality Explorer’s MDAO capability, it’s now possible to perform system-level optimization, from the IC to the package, the PCB and the system, in a fraction of the time.”