IRIS-HEP Fellow: Leonid Didukh
Fellowship dates: Jun – Sep, 2023
Project: Estimation of the energy cost and efficiency of the HEP data compression ML algorithm (Baler)With the increasing amount of information, the management and preservation of research data become memory and energy costly. Fast and cost-efficient software for data compression is necessary in order to resolve this problem. This can be achieved by Baler - neural compression software, but it is required to have the information and understanding of how the software allocates and consumes resource and energy. In this project, we will estimate the computational resource and energy consumption of the Baler framework.
More information: My project proposal
Caterina Doglioni (Lund University)