Netlist Simulation Speedup with Vanti
how to use explainable AI to speed up traditional net list based simulations
Net-list simulations are the backbone of semiconductor design and validation. They are the widely accepted - tried and proved - method to gain confidence in the product’s future.
But while simulators are the source of truth for meeting specifications, design goals, timing issues and more, they have remained largely unchanged and slow. It’s clear that any boost to the simulation’s throughput will dramatically increase an organization’s ability to achieve its design and validation goals faster and with reduced effort.
This is where Vanti steps in.
Vanti’s explainable AI can replace an entire net-list or parts of it with a closed-form analytic formula, negating the need for a full simulation. Using Vanti it is possible to reduce simulations from weeks to minutes, either increasing coverage or speeding up significantly the time it takes to achieve coverage.
This benefit extends to design as well, where the Monte-Carlo methodology is often the most common technique to explore design. Replacing parts or all of the design with an accurate closed-form analytic formula can reduce exponentially the number of different simulations to achieve coverage.