Tool released for accurate, hardware-validated CPU performance, power, and energy modelling

Full-system performance and energy simulations frameworks, such as gem5 and McPAT are heavily relied upon in research and development. The full methodology implemented in GemStone allows gem5 CPU models to be directly compared against hardware platforms and uses statistical and machine learning techniques to analyse and evaluate sources of error, without relying on detailed CPU specifications. This allows users to:

1. Evaluate and understand whether a specific gem5 model is fit for their intended purpose/use-case;
2. Make iterative improvements to gem5 CPU models and create new models for different platforms;
3. Validate gem5 models after changes have been made and check for consistency between different versions of gem5;
4. Use empirical power models to conduct energy analysis in gem5.

More details of our methodology can be found in the following publications:

M J. Walker, S. Diestelhorst, S. Bischoff, G. V. Merrett and B. Al-Hashimi Hardware-Validated CPU Performance and Energy Modelling in IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS), April 2018.

For more information, and to download the tool, visit www.gemstone.ecs.soton.ac.uk.