Parallel solvers comparison for an inverse problem in fractional calculus

High-Performance Computing (HPC) is a fundamental tool for improving the performance of many algorithms in terms of time, especially for large-scale problems. In the last years, various HPC architectures have been developed to quickly process data in many research areas and at the same time the HPC tools has become very important. In addition, the development of scientific libraries for parallel computing plays a key role in achieving better performance. In particular, thanks to the computational power of Graphic Processing Units, themost popular and inexpensive accelerators, the parallel computing field has become almost a standard process for datamanagement. Hence, the porting of many standard numerical libraries on these architectures produced excellent results. In this work, we deal with a two-dimensional time fractional diffusion problem. More in detail, we analyze the performance of some parallel codes, specifically designed to solve it, implemented in different architectures. Moreover, a further GPU version is proposed and compared with the above implementations.