Case Study: M31 Technology Corporation achieves 15x faster characterization and 77% higher accuracy for 12nm IP with Synopsys PrimeLib on Synopsys Cloud

M31 Achieves 15x Faster Characterization with 77% Higher Accuracy for 12nm IP on Synopsys Cloud with PrimeLib

M31 Technology, a Taiwan-based silicon IP provider specializing in high-speed interface IPs, memory compilers and standard cell libraries, needed to accurately characterize mixed-type 12nm standard cell libraries (combinational, sequential and MBFF) to achieve >90% first-pass silicon success. Their existing cloud deployment and tools could not meet the required accuracy, scalability or secure, user-friendly environment for large parallel characterization.

M31 adopted Synopsys PrimeLib on the Synopsys Cloud SaaS platform with FlexEDA per-minute licensing and PrimeSim HSPICE AVA, scaling to 8,000 CPUs for massive parallel runs and ML-based high-sigma characterization. The result: characterization completed 15× faster, passing rates improved by ~77% (with many library data types reaching up to 100% passing), Monte Carlo simulations finished in one day, and the work ran in a secure SOC 2 compliant cloud environment.

M31 Technology Corporation

Patrick Lien

Vice President of R&D