Result description
A set of heuristically-built power models based on hardware performance counters. Through their flexible API, the power models seamlessly integrate in other tools, such as AMPERE’s multi-criteria optimization, where they provide energy estimates for optimization of non-functional constraints. Runmeter also exploits the power models, providing reliable energy monitoring capabilities at runtime.
The usage of performance counters to estimate power consumption allows to address the heterogeneity and massive parallelism of AMPERE target platforms, also supporting DVFS. As such, despite their low overhead, they achieve high accuracy and responsiveness.
Addressing target audiences and expressing needs
- Business partners – SMEs, Entrepreneurs, Large Corporations
End-users, early adopters and new customers
- Public or private funding institutions
- Other Actors who can help us fulfil our market potential
- Research and Technology Organisations
R&D, Technology and Innovation aspects
Current: implemented for CPU adn GPU
Next: testing and validating on different others heterogenous architectures
Open Source via repository: Software publicly available in gitlab ( https://gitlab.bsc.es/ampere-sw/WP3/power-modeling )
Can clone from OS repository
Long-term maintenance via community

