Our group focuses its activity on advancing scientific and technological knowledge in two strategic domains: High-Performance Computing (HPC), as the foundation for advanced computational support, and the development of applications aimed at society, particularly in the field of Simulation and Management of Intelligent Healthcare Systems.
We work on expanding the capabilities of our HPC tools and developing new strategies to address Artificial Intelligence (AI)–based applications. Our objective is to design models and tools that enable safe, efficient, and energy-aware execution of applications on HPC systems, while also contributing to advancements in simulation and intelligent healthcare services.
Our research lines include the development and enhancement of key tools such as PAS2P and RADIC, their application to AI workloads, and the design of strategies that consider critical aspects such as performance, scalability, fault tolerance, security, and energy efficiency.
Within the area of Simulation and Management of Intelligent Healthcare Systems, we promote the development of technologies that allow the modeling, optimization, and strengthening of healthcare system resiliency. This includes the creation of simulators to support decision-making in Emergency Healthcare Services, high-performance simulation for chronic disease scenarios, and agent-based modeling of urban mobility during emergency situations.
Across all these areas, the need for optimal or near-optimal solutions to complex problems naturally arises. For this reason, we also investigate the design of efficient heuristic algorithms, applicable across a broad range of scientific and technical challenges.
Main Research Areas:
Efficient and Secure Execution on HPC Systems
-
Performance models for AI applications in HPC/Cloud: Extension and optimization of the PAS2P framework to improve efficiency through advanced mapping policies.
-
Efficient HPC I/O management for AI applications.
-
Integration of fault tolerance and security mechanisms into HPC systems.
Heuristic Algorithms for Complex Problems
-
Design of metaheuristic algorithms for search in n-dimensional spaces.
-
Application of these techniques to the scientific challenges addressed by the group.
Simulation and Management of Intelligent Healthcare Systems
-
Development of simulators for Emergency Departments (ED).
-
Automatic and Intelligent Management of Emergency Services (AIMED).
-
Resilience strategies for healthcare systems facing disruptive situations.
-
High-performance agent-based simulation for chronic kidney disease.
-
Agent-based simulation of urban mobility during emergency traffic scenarios.
Next figure shows the research lines structure:
Links of Interest:
https://ddd.uab.cat/collection/hpc4eas
https://portalrecerca.csuc.cat/sgr/2021SGR0128?locale=es
https://zenodo.org/communities/hpc4eas/records?q=&l=list&p=1&s=10&sort=newest