Mentor Graphics Corp., Wilsonville, Ore., has expanded the Kronos Cell Characterization and Analysis platform to include embedded memories. The platform quickly produces accurate performance models for standard cells, I/Os, complex cells, and embedded memories within an advanced, integrated environment. The Kronos Platform’s advanced algorithms and efficient job distribution reduces characterization time from weeks to days.
In managing the characterization and analysis of embedded memories, the Kronos platform delivers ease of configuration that enables an automatic characterization flow. The automated flow runs netlist reduction, all simulations and creates complete .libs. The flow employs fast simulation techniques and precise circuit setup and measurement controls.
Characterization and analysis of embedded memories poses unique challenges: large circuit size can lead to excessive runtime. Complex internal circuitry and a variety of choices available for pin-bus models require a high level of automation. The Kronos platform addresses these challenges with proven, integrated capabilities.
At 45nm and below, speed and power consumption are much more sensitive to environmental conditions, including voltage and noise. Therefore, producing accurate models at the appropriate conditions is critical to achieving design success. The Kronos platform quickly generates accurate and complete timing/power models and incorporates unique methods for noise immunity and signal integrity to avoid design problems that otherwise might not be detected until failure analysis.
