Uncategorized

MultiXscale expert, Dr. Rodrigo Bartolomeu and members of MultiXscale CoE, have received one of the two Best Papers Awards at the EuroHPC User Days 2026 for: “Scaling and performance aspects of the P3M method on multiple GPUs”. This work was selected by the scientific review board for its clear and strong contribution to making complex scientific simulations run faster and more efficiently on modern supercomputers. Abstract: Long-range electrostatic interactions constitute a major computational bottleneck in particle-based simulations, particularly when periodic boundary conditions are employed. Mesh-based Ewald methods, such as the particle-particle particle-mesh (P3M) approach, reduce the computational complexity to O(N log N) but introduce challenges related to parallel scalability, communication overhead, and efficient utilization of modern heterogeneous architectures. While highly optimized implementations exist in established molecular dynamics packages, these are typically tightly coupled to full simulation frameworks, limiting their flexibility and reuse in emerging HPC applications.In this work, we present a performance-portable library for electrostatic solvers, providing implementations of both classical Ewald and P3M methods targeting CPU and GPU architectures. The library is designed to enable efficient execution on multi-GPU systems while maintaining portability across heterogeneous platforms. We investigate scaling and performance characteristics of the P3M method, with particular focus on the FFT-based long-range component, which is known to limit scalability. Strong-scaling experiments demonstrate that small problem sizes are dominated by communication overhead, leading to reduced parallel efficiency, whereas larger systems achieve near-ideal scaling over a wide range of GPU counts.Our results highlight the importance of balancing computation and communication in mesh-based electrostatics and demonstrate that portable, decoupled implementations can achieve competitive performance on modern HPC systems. The presented approach facilitates integration into diverse simulation workflows and provides a foundation for further optimization of long-range solvers in exascale computing environments. More information available at EuroHPC JU website here

MultiXscale CoE will collaborate with the activity “Mini-supercomputers in your hands made with LEGO bricks”, organized by the Computing and Advanced Technologies Foundation of Extremadura (COMPUTAEX) on 25 September from 17h30 to 21h, at Espacio UEx Cáceres (Spain). This hands-on workshop is designed for children aged 6 to 12, where they will learn about how to build small supercomputers using LEGO bricks and Raspberry Pi boards. These compact systems, roughly the size of a credit card, combine low energy consumption with an affordable cost. Participants will discover each of the system’s components and the connections required to make them work. They will also learn what a supercomputer is and understand the advantages it offers compared with a conventional home computer. In addition, they will explore how scientists and researchers across a wide range of fields often rely on supercomputers to tackle important and complex problems that would otherwise be impossible to solve. More information available here

The following articles, published in Procedia Computer Science by MultiXscale experts, will be presented at the Parallel Session “Chemistry and Material Sciences” organized on Wednesday, 23 September from 14h to 15h30, in the framework of the EuroHPC User Days Dublin 2026:– Scaling and performance aspects of the P3M method on multiple GPUs– Coupling LAMMPS with LPC3D for Mesoscopic Simulations of Electrolytes in Porous Carbons: A Python-MPI Workflow for CPU and GPU Supercomputers – Coupling ESPResSo and waLBerla for scalable soft matter simulations with hydrodynamic interactions Besides, Alan O’Cais, EESSI Coordinator, will participate in the Plenary Session “EuroHPC User Forum”, on Tuesday, 22 September from 14h to 15h30. Further information about the event is available here

New paper available at AIP Publishing, Physics of Fluids 38, 088104 (2026): Generalized Onsager-regularized lattice Boltzmann framework by Anirudh Jonnalagadda et al. https://doi.org/10.1063/5.0327546 Abstract: This work presents a novel strategy to address Navier–Stokes modeling errors arising in first-nearest neighbor lattice Boltzmann methods and introduces fully local corrections through Onsager-regularized (OReg) non-equilibrium populations. The proposed mechanism, which admits partially and completely corrected OReg models, is used to develop representative partially and completely corrected models for the six-moment-constrained guided equilibrium (GEq) representation on the D2Q9 lattice. The former realization only addresses compatibility condition violations and improves the accuracy by two/four orders of magnitude at reference/arbitrary lattice temperatures, respectively, while the latter additionally corrects stress tensor modeling errors, resulting in a fully corrected exact model. Numerical benchmarks of the corrected schemes demonstrate improved accuracy and stability in comparison to the lattice-BGK, regularized, and uncorrected OReg-GEq schemes, thus presenting a promising avenue for OReg based thermohydrodynamic extensions.

Are you interested in scientific software, HPC, and reproducible computing environments? Join us in Dublin for an exciting EESSI Hackathon on Friday, 25 September 2026, from 09:00 to 17:30 local time (IST, UTC+1). Taking place immediately after the EuroHPC User Days, held in Dublin from 22 to 24 September, the hackathon will offer a great opportunity to get hands-on with EESSI, exchange expertise, tackle challenges, and collaborate with fellow members of the HPC community. Whether you are already using EESSI or are keen to discover how it can simplify the deployment and sharing of scientific software across diverse computing environments, this is a great opportunity to learn, experiment, and contribute to the EESSI ecosystem. Don’t miss this opportunity to connect with the EESSI community and take your HPC software workflows to the next level! Provided by EuroHPC CoE MultiXscale in collaboration with the EuroHPC Federation Platform. Further information and registration available here.