Linux Systems · HPC · C++/CUDA · Numerical Methods
Paolo Joseph Baioni
Linux systems administrator at CERN, in the Infrastructure, System Administration & AI (ISA) section of the Controls Software and Services (CSS) group, Beams (BE) department, working on the computing platforms behind the CERN accelerator control system.
Previously Technical Manager at the MOX Laboratory (Politecnico di Milano), leading HPC cluster design and deployment and developing GPU-accelerated scientific software, with a focus on C++/CUDA development, numerical methods, and performance portability across heterogeneous CPU/GPU architectures.
Background in computational physics and numerical methods, with experience in large-scale simulations and performance-critical applications. Contributions include peer-reviewed publications and the development of scientific software.
Experience
2026–Present
Member of the Linux system administration team for the accelerator control system, within the Infrastructure, System Administration & AI (ISA) section of the Controls Software and Services (CSS) group.
Contributing to the evolution of the core platform through the implementation of new storage services based on a high-end commercial storage appliance, including its integration into the existing configuration management systems, access control and data segmentation, redundancy and backup schemes, and the migration of control system data from existing NFS servers.
2023–2026
Led HPC infrastructure development and scientific computing strategy. Designed GPU-enabled systems and developed high-performance scientific software, while coordinating technical activities and supporting research projects.
Work included development of scientific software from scratch, later adopted and extended in subsequent research projects, as well as analysis and extension of existing codebases and adaptation of HPC software stacks (e.g. OpenPBS) to heterogeneous Linux environments.
Co-supervision of MSc theses and of a PhD student on HPC, GPU computing, and parallel programming topics.
2022–2026
Delivered MSc-level laboratories in scientific computing (C++, MPI, GPU programming) and in numerical modelling and methods, across several courses.
2021–2025
Developed massively parallel numerical methods and GPU-accelerated simulation codes with performance portability across architectures, alongside dissemination activities including publications, conference presentations, and organisation of minisymposia.
Technical Background
Experience in developing and optimising numerical methods for large-scale simulations and scientific applications.
Focus on GPU optimisation, parallel algorithms, and development of portable, high-performance scientific software for research applications.
Research & Selected Projects
Implementation of a massively parallel Material Point Method in C++20 using TBB and (roc)Thrust, targeting CPU/GPU architectures with performance portability.
Adopted for continued development within an ongoing PhD project.
Design and deployment of modern GPU-based HPC systems to support research workloads.
Plasma edge modelling for nuclear fusion using SOLPS-ITER. Background strengthened through participation in the Max Planck Institute for Plasma Physics Summer School.
Implementation of a neural network from scratch in C++ using Eigen and STL.
Repository →Links
Contacts
paolojoseph dot baioni at cern dot ch
paolojoseph dot baioni at polimi dot it
paolojoseph dot baioni at gmail dot com