Gem5 is a widely-used architectural simulator that models a diverse set of parameters. We explored the gem5 design space and modelled application-class SoCs like CVA6 core model on the gem5 platform. The model parameter optimization was carried out using Machine Learning techniques, while its performance was assessed using MiBench and RISC-V micro-benchmarks, in both the SE and FS modes of gem5. Currently, we are automizing the SoC modelling process on gem5 to reduce the time and complexity of the process.
