Our Group integrates a series of lines of research, interrelated and developed in the context of High Performance Computing (HPC)

Research Lines

The HPC4EAS research group focuses on High-Performance Computing (HPC), efficient and secure computing, heuristic methods, and high-performance simulation, with particular attention to Artificial Intelligence applications and intelligent healthcare systems.

Our research is organized into three main areas:

1. Efficient and Secure Execution in HPC

Research on methodologies, models, and tools to improve the efficient and secure execution of applications on HPC and Cloud Computing infrastructures, with particular emphasis on Artificial Intelligence workloads.

  • Performance Modeling of AI Applications in HPC and Cloud Computing
    • Performance analysis and modeling of AI applications.
    • Performance prediction and scalability.
    • Efficient mapping of applications onto computing resources.
    • Computational and energy efficiency.
  • Efficient HPC I/O Management for AI Applications
    • Characterization and analysis of I/O behavior in AI workloads.
    • Efficient management of data access and movement.
    • Performance and scalability of HPC storage and I/O systems.
    • Strategies for improving the execution efficiency of data-intensive AI applications.
  • Fault Tolerance and Security in HPC Systems
    • Fault-tolerance mechanisms for HPC environments.
    • Integration of fault tolerance and security.
    • Software vulnerability analysis and attack detection.
    • Resilience and secure execution of HPC applications.

2. Heuristic Algorithms for Complex Problems

Research on computationally efficient heuristic and metaheuristic methods for finding optimal, near-optimal, or high-quality solutions to complex problems.

  • Metaheuristic Search in Multidimensional Spaces
    • Design and evaluation of metaheuristic search algorithms.
    • Exploration of large and complex solution spaces.
    • Computationally efficient search strategies.
  • Application of Heuristic Methods
    • Resource allocation and configuration in HPC systems.
    • Intelligent management and decision-support problems.
    • Application of heuristic approaches across the group’s HPC and healthcare research activities.

3. Simulation and Management of Intelligent Healthcare Systems

Research on modeling, simulation, and intelligent decision-support systems for healthcare applications, combining HPC, Artificial Intelligence, and Agent-Based Modeling (ABM).

Evaluation of mobility scenarios affecting emergency services.

Emergency Department Simulation

-Modeling and simulation of hospital Emergency Departments.

-Development of configurable and on-demand simulators.

-Simulation-based decision support.

Automatic and Intelligent Emergency Department Management – AIMED

-Intelligent management of healthcare resources.

-Decision-support systems for Emergency Departments.

-Digital Twin approaches to anticipate system behavior and critical situations.

Resilience of Emergency Departments

-Analysis of healthcare services under disruptive and extreme situations.

-Strategies to maintain service performance under overload conditions.

-Simulation of accidents, catastrophes, pandemics, and other critical scenarios.

High-Performance Agent-Based Simulation for Chronic Diseases

-High-performance Agent-Based Modeling and Simulation.

-Modeling the progression of chronic diseases.

-Support for healthcare decision-making, with particular attention to Chronic Kidney Disease.

Agent-Based Simulation of Urban Mobility and Emergency Traffic

-Modeling urban mobility using Agent-Based Simulation.

-Analysis of high-priority emergency traffic.

Contact us

News

New CORA Dataset on Parallel I/O in Distributed Deep Learning

Edixon Parraga and collaborators have made a new research dataset openly available through CORA – Research Data Repository.

Replication Data for: Parallel I/O Analysis in Distributed Deep Learning Applications on High-Performance Computing

The dataset contains experimental data from the study of parallel I/O in distributed Deep Learning workloads on HPC systems. Its publication provides the scientific community with access to the data behind the research, supporting further analysis, comparison, and reproducibility of the results.

Access the dataset:
DOI: 10.34810/DATA3618

New Research Datasets Published in CORA

Two new research datasets by Maria Harita Rascón and collaborators have been published in CORA – Research Data Repository, reinforcing our commitment to open science and reproducible research.

The datasets provide replication data supporting research on sampling- and clustering-based heuristic search algorithms and their application to optimization problems:

Replication Data for: A Metaheuristic Search Algorithm Based on Sampling and Clustering
DOI: 10.34810/DATA3157

Replication Data for: Application of a sampling and clustering-based heuristic search algorithm to find an efficient staff configuration in an emergency department
DOI: 10.34810/DATA2951

These publications make the experimental data openly available to the research community, facilitating the validation, reuse, and reproducibility of the research results.

News Blog

More blog entries