SURF contributes to NL-ECO to investigate concepts for energy efficient information technology
Energy-efficiency in scientific computing is one of the prominent topics of innovation within SURF. In our SURF Open Innovation Lab, we are assessing various technologies and approaches through experimentation and collaboration. To optimise balancing performance of scientific workloads vs. energy consumption levels, including energy management frameworks on the supercomputer as well as different hardware components, for example.
Part of our interest in this NWO Granted proposal is the translation of innovation emanating from NL-ECO into new possibilities for supporting scientific research, just in line with our innovation objectives and core mission.
Our activities in NL-ECO: investigating feasibility of new technologies
SURF wishes to focus on and contribute to value creation, knowledge building and evaluation of usability and feasibility aspects of researched technologies in this consortium. We will facilitate and participate in designing prototype experiments and conduct pilot studies together with other partners.
Specifically, SURF in collaboration with Radboud University, plans to investigate the potential and quantify the efficiency of neuromorphic coprocessors for optimisation and machine learning tasks in scientific computing. In collaboration with industry, we aim to assess the energy cost of computations using the experimental platform at SURF.
Contribution to our national IT infrastructure
We expect that the utilisation of the obtained knowledge and developments from this project could be applied to our national IT infrastructure for research and education.
Do you want to read more about this consortium? Read the news release ‘Big consortium starts research into energy-efficient information technology’ on the website of Utwente