cfaed Publications
Energy Efficiency for edge/cloud optimization
Reference
Jeronimo Castrillon, "Energy Efficiency for edge/cloud optimization", In EMEC Workshop on Energy and Material Efficiency in Cloud-Edge continuum, co-located with the international conference on Design, Automation and Test in Europe Conference (DATE) (invited talk), Apr 2026.
Abstract
This talk presents the MYRTUS Design and Programming Environment (DPE), an interoperable framework for modeling and deploying applications across the computing continuum. The environment features an MLIR-based infrastructure to support diverse hardware targets while allowing the seamless import of third-party code. We report on current support for FPGA and CGRA targets. A key aspect of the DPE is to support adaptability in the computing continuum by exploring operating points at design time, enabling a runtime engine to dynamically adapt configurations for energy efficiency. Additionally, the talk covers reactive programming, including enclaves and quasi-static scheduling to bridge the gap between high-level models and low-level executable schedules, ensuring efficient and deterministic execution in the context of cyber-physical systems.
Bibtex
author = {Castrillon, Jeronimo},
howpublished = {EMEC Workshop on Energy and Material Efficiency in Cloud-Edge continuum, co-located with the international conference on Design, Automation and Test in Europe Conference (DATE) (invited talk)},
month = apr,
title = {Energy Efficiency for edge/cloud optimization},
year = {2026},
abstract = {This talk presents the MYRTUS Design and Programming Environment (DPE), an interoperable framework for modeling and deploying applications across the computing continuum. The environment features an MLIR-based infrastructure to support diverse hardware targets while allowing the seamless import of third-party code. We report on current support for FPGA and CGRA targets. A key aspect of the DPE is to support adaptability in the computing continuum by exploring operating points at design time, enabling a runtime engine to dynamically adapt configurations for energy efficiency. Additionally, the talk covers reactive programming, including enclaves and quasi-static scheduling to bridge the gap between high-level models and low-level executable schedules, ensuring efficient and deterministic execution in the context of cyber-physical systems. },
date = {2026-04},
location = {Verona, Italy},
}
Downloads
No Downloads available for this publication
Permalink
https://cfaed.tu-dresden.de/publications?pubId=3888


