AI-Accelerated Radiative Transfer Simulations for SKAO (ARTS4SKA)
Description
This project proposes the development of an advanced radiative transfer code for large-scale simulations of the Cosmic Dawn and Epoch of Reionization, optimized for upcoming SKA-Low observations. A key innovation is the integration of AI-driven methods into the chemistry solver to significantly accelerate and enhance the accuracy of convergence in modeling the complex evolution of primordial gas.
Combined with GPU-accelerated raytracing and a new particle-to-mesh framework, this AI-enhanced approach will enable physically detailed simulations across cosmological volumes.
Key Data
Projectlead
Deputy Projectlead
Project partners
Eidgenössische Technische Hochschule Zürich ETH; Universität Basel; Ecole polytechnique fédérale de Lausanne EPFL; Swiss National Supercomputing Centre (CSCS)
Project status
ongoing, started 01/2025
Institute/Centre
Institute of Business Information Technology (IWI); Centre for Artificial Intelligence (CAI)
Funding partner
Swiss National Supercomputing Centre (CSCS)
Project budget
95'033 CHF
