Xcelium ML for 5X Faster Regression Throughput

Overview

Regressions time often becomes one of the biggest challenges to meet the tight project schedule with increasing complexity of the SoC designs and shorter time to market. Verification engineers apply a coverage-driven methodology and run a large number of constrained random tests with multiple seeds in massive regressions to meet their coverage goals. Thus, verification closure is a long and tedious job and a great area to apply machine learning.

Cadence® Xcelium Logic Simulator is the industry’s first simulator to leverage powerful machine learning technology (ML) to deliver up to 5X faster regression throughput. Xcelium ML learns iteratively over an entire simulation regression to observe the different randomization points and creates a machine learning model that correlates the randomization and coverage. It analyzes the patterns hidden in the verification environment and intelligently generates an optimized regression achieving the same coverage significantly faster. In-addition, it generates analytics on random knobs and coverage bins, and provides other interesting use cases to meet user-directed verification goals such as maximum throughput, stress design instances, generating a regression focused on specific areas of the design (coverage space) and replacing directed tests with randomized tests etc.

Join this webinar to learn how Xcelium ML can be a game changer for your verification flow, providing up to 5X regression throughput saving significant time and compute resources.

Date and Time

Thursday, October 27, 2021
10:00 AM (PT) / 1:00 PM (ET)