Cadence Design Systems recently announced that Arm will now use the Cadence Liberate MX Trio Characterization to improve and improve the quality of its embedded memory and compilers. With the Liberate MX Trio Characterisation, Arm achieved better accuracy and capacity requirements and reduced the execution time of Memory Freedom Variable Format (LVF) characterization by up to 7x compared to brute-force Monte Carlo simulations.
“Arm uses the Liberate Trio Characterization Suite for standard cell characterization, so it was an easy choice for us to extend our Liberate MX Trio Characterization deployment to meet our new LVF memory characterization needs,” said Philip Mayer, VP of Design Enablement. , Physical Design Group, Arm. “By incorporating the Liberate MX Trio Characterization into our methodology, we are improving accuracy, capacity and achieving our go-to-market goals with the delivery of our embedded memory instances and compilers.”
“Memory characteristics can have a significant impact on signing accuracy, and Arm was looking for a robust solution for their embedded memory IP that would help them meet accuracy requirements while accelerating time to market,” said Sharad Mehrotra, VP of Research. and Development in the Digital & Signoff Group at Cadence. “Arm has joined the community of successful, production-proven Liberate MX Trio Characterization customers, trusting the solution for their memory characterization needs and expanding their use of Liberate’s broader product portfolio for standard cell power and performance characterization.”
Liberate MX Trio characterization is a library characterization solution designed for large, advanced node memory designs, enabling Arm engineers to perform on-board memory and compiler characterization checks efficiently and accurately. Liberate MX Trio Characterization provided Arm with automated and simulation-based dynamic resolution to reduce runtime, increase capacity, and maintain SPICE-level performance and accuracy. Modeling on the full list of RC connections, spanning the full vector set, allows Arm to identify the exact worst-case paths and maintain SPICE-level accuracy.
Support for the new Cadence library for Arm
Source link Support for the new Cadence library for Arm