Since my incorporation to the Department of Computer Architecture and Operating Systems of the UAB, I have developed all my research activity in different aspects related to parallel processing:

Automatic analysis and dynamic tuning (1999 -):

This line focuses on automatic performance analysis and dynamic tuning of parallel and distributed systems. In this approach, an application is monitored, its performance problems are detected, solutions are given and finally it is tuned during the execution time. We have developed several performance tools: MATE (Monitoring, Analysis and Tuning Environment), GMATE (MATE for grid environments), ELASTIC (performance tuning for large-scale environments). This research is closely related to the parallel application modeling (definition of performance models for applications developed for various platforms). Considering the benefit of both lines and their relationship, the development of other models has begun especially agent-based models for simulations of different real-based situations systems. A progress has been done in optimization of the distribution of agents according to their computational load and communication patterns, their integration into automatic analysis and dynamic tuning environments, as well as designing and implementing of benchmark to evaluate different agent-based platforms. The use of machine learning techniques is also proposed to analyze and tune applications based on different paradigms (OpenMP, MPI, accelerators). In this sense, there is a preliminar work that opens the door for developing scalable and efficient methodologies capable of adapting autonomously and intelligently to the increasing complexity of applications and hardware.

In 2011 the European project FP7 "AutoTune: Automatic Online Tuning" (2011-2014, Ref: 288038, web: http://www.autotune-project.eu/) was financed jointly with TUM Germany, CAPS France, UNIVIE Austria, ICHEC Ireland, LRZ Germany. In this project I participated as main researcher (on behalf of the UAB). The work was focused on the extension of the Periscope Tool (TUM) and our group supported the tuning of parallel applications and analysis of typical performance problems for MPI applications. Given the good results obtained, the collaboration between the partners has been maintained to continue advancing in the autotuning issue. 

In order to continue researching the issues of autotuning and agent-based simulations, collaboration has been extended with the BSC (Barcelona), RCSI (Ireland), Oregon University and Iowa University.

Throughout these years, my participation in reviews of several journals with a high impact factor and in international congress committees has increased. I have been also chosen as an expert-evaluator for different calls of H2020 projects.

National projects: 9, European: 3

PhD Thesis presented in 2004

Advisor of: Master Thesis: 6, PhD Thesis: 4 (2 in progress)

Articles in indexed journals: 16, non-indexed: 2

International conferences: 30, national: 7

Other publications: 5, International invited presentations: 9

After my thesis defense, I participated in the European project called Crossgrid whose objective was to develop efficient management policies for parallel and interactive applications. When I joined the project, I was assigned the topic related to the management of workflows in Grid environments. In my work I presented a solution on how to describe a workflow in the form of a DAG, and how to carry out its execution and control afterwards.

European project: 1, International Conferences: 1

Finally, to mention my personal and technical experience, I participated in a cooperation and technological development agreement that was established with the company Oryzon Genomics financed with a PROFIT project and related to biotechnology (design of DNA chips by comparative genomic hybridization and ultra-quick genetic mapping). My work consisted of studying the requirements of the existing application for its parallelization in a cluster environment. The developed software was validated by preparing the environment and the application conducting both practical tests and massive executions.

Number of six-year terms granted: 2; Last term: 2009-2014.

Accreditation by ANECA for the associated professor category (2015)

Number of Doctoral Theses supervised in the last 10 years: 2 (2 in progress) 

- Claudia Rosas. Performance Improvement Methodology based on Divisible Load Theory for Data-Intensive Applications. Start: 27/11/2009 End: 16/07/2012. Number of conferences: 2. Number of journals: 2. 

- Andrea Martínez Trujillo. Dynamic Tuning for Large-Scale Parallel Applications. Start: 05/11/2010. End: 09/10/2013. Number of conferences: 4. Number of journals: 2. European mention. 

Number of Master Thesis conducted in the last 10 years: 5 

Total number of citations: Web of Science (Thomson): 165, Scopus: 253, Google Scholar: 477 

Average citations per year (2016-2020): Web of Science (Thomson): 15, Scopus: 24.4, Google: 37.6 

Total number of publications in 1st quartile: 2, 2nd quartile: 10, 3rd quartile: 2, 4th quartile: 2  

H-Index: Web of Science (Thomson): 7, Scopus: 8, Google Scholar: 14 

Expert evaluator of applications for research projects of the Horizon2020 program

- Work Program 2016-2017, ICT-05-2017

- FET Work Program 2018-2020, FETOPEN-01-2018-2019-2020: FET-Open Challenging Current Thinking

External expert for monitoring the projects founded by the H2020 calls:

- H2020-FETHPC-2017

- H2020-SME INSTRUMENT

Reviewer of journals: Journal of Parallel and Distributed Computing (JPDC), Future Generation Computer Systems (FGCS), Simulation Modeling Practice and Theory, Journal of Supercomputing, IEEE's Transactions on Services Computing (TSC), Journal of High Performance Computing (JHPCA), Journal of Parallel Computing, Journal of Computation Science

Reviewer of international conferences: ICCS, EuroPar, ParCo, CCGrid, HPCC, IPDPS, ICPP, PPAM, EuroMPI, PACT, AINA, ICA3PP, ISAT, SC, SC2

Member of the Program Committee: EuroPAR 2009, 2015, ICPP 2009, 2021 PPAM2011, 2013, 2015, AINA 2013, 2014, EuroMPI 2012, 2013, 2014, 2015, ISAT 2015, 2016, 2017, 2019, ICCS 2016, 2020, ICA3PP 2016, SC2 2016, 2017, 2019, SOCA 2017, 2018, SEM4HPC 2017, HPDC 2020, 2021

Co-organizer of the mini-symposium Application Autotuning for HPC (Architectures) of the ParCo 2013 conference.