Coordinator
Prof. Luca Heltai
e-mail: luca.heltai@unipi.it
Administrative seat
Department of Mathematics
Project description
The PhD programme in “High Performance Scientific Computing” (HPSC) has been designed to train a new generation of highly qualified researchers in the field of high-performance scientific computing and its multidisciplinary applications. This industrial, interdisciplinary and interdepartmental programme is organised in collaboration between the University of Pisa and the company S.I.T. – Sordina IORT Technologies S.p.A.
The HPSC PhD programme promotes the development of innovative computational methods across various scientific and application areas, employing High-Performance Computing (HPC) to build innovative solutions and address complex problems in mathematics, computer science, engineering, physics, earth and climate sciences, chemistry, and life and health sciences. Through multisectoral collaboration between the various departments and with the industrial partner, the programme is designed to train a new generation of researchers capable of pushing the frontiers of multidisciplinary research and driving technological innovation to tackle real-world challenges through high-performance computing.
HPSC strongly encourages the development of interdisciplinary research projects involving many departments that focus on applying High-Performance Computing (HPC) to address complex, cross-cutting challenges. Each PhD student will acquire advanced, field-specific expertise in HPC, including algorithms, software, and artificial intelligence techniques applicable across different contexts. In addition, PhD students will have the opportunity to work on a concrete application project, selected from strategic areas such as climate modelling, biomedical research, and smart city technologies, to maximise the impact of the results in fields of high scientific and social relevance.
The HPSC programme includes:
- A structured educational offer with foundational, methodological and thematic courses developed jointly by the departments involved;
- Participation in interdisciplinary and industrial research projects;
- Integration into international networks and promotion of international mobility;
- The development of high-impact products such as open-source software, advanced algorithms, patents and innovative solutions for research and industry.
Particular emphasis is placed on the responsible and efficient use of computational resources, promoting sustainable practices in scientific computing. The training programme also addresses issues related to the energy consumption of HPC systems and the environmental impact of large computing infrastructures, encouraging the adoption of technical and algorithmic solutions aimed at reducing ecological footprint. Furthermore, the integration of artificial intelligence is approached from a critical and informed perspective, with the aim of developing transparent, efficient and sustainable approaches.
The scientific and strategic relevance of the HPSC training project is further underlined by the fact that the Department of Mathematics at the University of Pisa has been selected as one of the European Centres of Excellence within the EuroHPC JU (European High Performance Computing Joint Undertaking) initiative. This official recognition confirms the University’s international leadership in the HPC field and further consolidates the value and impact of the HPSC PhD programme within the European advanced research landscape.
The HPSC programme aims to provide PhD students with a solid theoretical and technical foundation in scientific computing, with a particular focus on the use of high-performance infrastructures to address complex problems across a wide range of scientific domains.
Computational Mathematics and Algorithm Development: Development of advanced mathematical models and optimised algorithms for HPC environments. This includes numerical methods for parallel computing, optimisation techniques, and simulation methods applicable in various fields, drawing on the expertise of the Departments of Mathematics and Computer Science in numerical linear algebra, finite element methods, reduced models, optimisation, etc.
HPC Software and Systems: Development of scalable, efficient and robust HPC software systems. Research in this area involves improving supercomputer architecture, parallel computing, and operating systems to enhance computational efficiency and energy sustainability, drawing on the expertise of the Department of Computer Science.
Data Science and Big Data Analytics: Development of new solutions and HPC methods for managing, processing and analysing large datasets in scientific research. This includes the development of new parallel algorithms for machine learning, statistical methods, and data visualisation techniques to extract insights from complex datasets across a variety of domains.
Computational Engineering: Application of HPC to the resolution of complex engineering problems, including, for example, fluid dynamics, optimisation of energy networks and devices, materials science, structural analysis, and biomedical engineering. This area supports the development of new simulation methodologies that can lead to innovation in biomedical, aerospace, civil and energy technologies.
Physics and Quantum Computing: Use of HPC and distributed computing systems for the analysis of large datasets (exa-bytes) in experimental and observational physics. Development of computing accelerators in heterogeneous systems based on both GPUs and FPGAs, and their application in real-time contexts. Development of AI/ML techniques for simulation and data analysis in physics. Use of HPC infrastructures (CPU and/or GPU) and development of parallel algorithms for the simulation of systems of physical interest (e.g. Lattice QCD, fluid dynamics, and plasmas). Development of Quantum Computing algorithms and investigation of the interface between HPC and QC.
Chemistry and Molecular Modelling: Exploration of molecular modelling in classical, quantum, or multiscale frameworks and its applications within HPC computing. Use of machine learning methods to enhance classical and quantum approaches in modelling properties and processes of complex molecular systems. These topics aim to foster the application of HPC computing to chemical research and materials science.
Earth Sciences: Simulation, processing and probabilistic inversion of 2D, 3D and 4D geophysical data, also using AI techniques; HPC for satellite SAR interferometry: parallel algorithms for SAR data processing, multi-temporal methods and real-world physics modelling; numerical simulations for the study of climate change, using both measured data and climate proxies; mathematical modelling of aquifer systems for water sustainability; modelling of magma propagation and/or accumulation and magmatic fluids in the subsoil for volcanic risk mitigation; geodynamic simulations for studying the evolution of the lithosphere on long timescales.
Life and Health Sciences: Use of HPC methods for managing, processing and analysing datasets of biological and pharmacological relevance. Possible topics include molecular kinetics simulations of protein–protein interactions, modelling tools for cell functioning, cell signalling and gene enrichment, identification of disease biomarkers using predictive tools. HPC methods will also support drug design and drug discovery through the development of predictive molecular models for the identification of new potential drugs. In silico tools will also be applied to the design of clinical trials and the assessment of their results.
Medical Physics: Development of advanced computational tools to optimise and improve diagnostic and therapeutic technologies. Possible research areas include radiotherapy, with particular focus on Flash radiotherapy and multiscale radiobiological modelling, medical imaging and radiation dosimetry. Research activities include Monte Carlo simulations of beam transport, modelling of biological responses to treatments, and the use of artificial intelligence techniques to predict therapeutic effects and optimise treatment protocols. The integration of numerical methods and machine learning enables improved treatment planning and a deeper understanding of the mechanisms behind the effectiveness of advanced radiotherapy.
Scientific-disciplinary areas
Computational Mathematics and Algorithm Development
HPC Software and Systems
Data Science and Big Data Analytics
Computational Engineering
Physics and Quantum Computing
Chemistry and Molecular Modelling
Earth Sciences
Life and Health Sciences
Medical Physics
Website
https://www.dm.unipi.it/phd-hpsc/
