Interdisciplinary Foundations for Large-Scale Data and Simulation Science
Data structures of immense size, variety, and complexity are transforming scientific discovery across disciplines. To extract meaningful insights, methods and tools that go well beyond classical statistics and serial computing are essential. Building, deploying, and validating such novel solutions for processing, visualization, and analysis of data, powered by high-performance computing systems, is a key area of expertise of the CDS. As models grow more complex, custom algorithms and software are no longer optional, they are essential for efficient, accurate simulations and machine learning.
The UoC Center for Data and Simulation Science (CDS) has been founded in January 2018 as a Central Research Institute.
The CDS bridges the gap between domain scientists and computational experts and is built on an interdisciplinary foundation that brings together three core domains: Data Science, Scientific Computing, and Domain Sciences. These areas converge to create a unified, collaborative research environment. Each domain represents a distinct yet interconnected research field:
- Data Science focuses on extracting insight from complex datasets.
- Scientific Computing provides the computational framework and advanced algorithms required for the processing of data, simulation of large-scale phenomena, and scalable and numerically efficient machine learning.
- Domain Sciences encompass application areas such as natural sciences, life sciences, economics, and the humanities, where these methods and technologies are applied to address real-world challenges.
The interaction between these domains gives rise to specialized areas of synergy. High-Performance Computing (HPC) & Scientific Machine Learning (SciML) connect Data Science with Scientific Computing, combining advanced analytics with high-performance computational methods. HPC and Computational Sciences link Scientific Computing and Domain Sciences, enabling sophisticated modeling and simulation within specific research areas. SciML and Research Data Management bridge Data Science and Domain Sciences by supporting domain-aware and -tailored machine learning and data analysis. At the core of this structure is a fully integrated, cross-disciplinary research environment in which domain expertise, computational power, and advanced data methodologies work together.
The unique strength of the CDS lies in a proven track record of excellent research which is combined with deep theoretical and technological foundations enabling innovation in data-driven and simulation science. It lies not in any single discipline alone, but in the dynamic collaboration across fields, supported by high-performance computing and tailored algorithmic solutions.