Energy wastage due to unnecessary recalculations
Algorithms are used for many expensive optimization problems, like designing a highly efficient next-generation solar cell for the project HELIOS. To find the best solution, specific computational techniques (optimisation methods) are developed. Before these are applied in practice, they must first be thoroughly tested. This testing consumes a great deal of energy, as thousands of computationally intensive simulations are run on high-performance computer clusters. As scientific research moves at a high pace on a global scale, identical simulations are frequently performed from scratch by different teams. This lack of coordination leads to repeated calculations, wasting precious time and energy.
The RECAB project, led by Vanessa Volz and Koen van der Blom, will develop a new open-source tool that enables the reuse of previous calculations by other researchers and stores new calculations for future use. This way, RECAB will drastically reduce energy consumption.
Democratisation of science
Another benefit is that computational results will also become available to researchers with smaller budgets or those with limited access to supercomputers. This makes it possible for more research teams to conduct high-quality research into complex problems. Furthermore, it enhances the reliability of research, as researchers can now rely on robust and proven simulations, thereby preventing unintended biases in algorithms.
Collaboration and the future
The project is supported by an international consortium of experts from countries including the United Kingdom, Germany, Slovenia and Portugal, that will provide feedback to ensure usability for the wider research community. The software developed will be fully open source and the database will be hosted by CWI’s IT department to ensure long-term availability.
See all the funded projects on the NWO-website.
