See my list of publications:
Research Engineer at Hewlett-Packard Laboratories (Bristol, UK)
- Carried out research on controllability and observability properties of OpenNebula and OpenStack Cloud technologies.
- Analysis of the performance of hardware counter-based profiling for large infrastructures, aimed at developing algorithms to reduce the risk of Virtual Machines overload and data starvation.
- Research on security at hypervisor level in order to avoid inferences of the underlying physical system or neighbour VMs from a compromised VM.
- Systems Engineer for the HP Cloud infrastructure federated to the EU FP7 BonFIRE project. My tasks involved the design and build of the cloud site, as well as its administration on a daily basis, extending its capabilities and participating in the decisions taken at the BonFIRE consortium.
- Management of small groups (up to 2 people) for software development of research prototypes.
Researcher and developer at the University of Santiago de Compostela (Spain)
Project: “Using Hardware Counters to Improve Memory Performance: Irregular Codes And Page Migration” in collaboration with Hewlett-Packard Laboratories USA.
- This 3-year project investigated the use of hardware
counters on the Itanium platform (formerly IA-64) to perform run-time
optimisations, focused on the Event Address Registers (EAR) and the
- Development of new page migration algorithms based on performance hardware counters to optimise the execution of parallel scientific codes in run-time.
- Development of efficient strategies to optimise the execution of parallel (regular and irregular) codes in run-time on shared-memory multiprocessors.
- Study of the optimal thread-to-core allocation on NUMA systems using the latency access information provided by the performance counters.
- Additional duties involved the administration of a rx7640 node of the FinisTerrae supercomputer at CESGA (Galicia Supercomputing Centre), one of the largest shared-memory supercomputers in Europe.
Researcher at the University of Santiago de Compostela (Spain)
Project: “Performance analysis of parallel applications”.
- Behavioural study of different molecular dynamic codes such as GROMACS and LAMMPS.
- Development of parallelisation techniques for the sparse kernels of such codes focused on improving the memory and cache utilisation on NUMA systems.
- Involved in the Formiga Cloud project, which aims to reuse the computational resources of the university and other centres by deploying a virtualised platform using VMWare and VirtualBox.
- Development of a performance model based on the Roofline Model for the FinisTerrae supercomputer.
Software Developer at CEDETEL (Valladolid, Spain).
CEDETEL is the Centre for the Development of Telecommunications in Valladolid
- Design, programming and deployment of a distributed J2EE-based videoconference platform “Web Call Center” for a Spanish Telecommunication firm (RETECAL).
- Rational Unified Process analyst and designer.
- Application Server administrator: Bea Weblogic, Orion Server and J2EE Server.
- Systems administrator of a Red-Hat based server infrastructure.
Participation in R+D Projects
I have been involved in 16 research projects: 1 project at the University of Santiago de Compostela, 1 project at a private company (Hewlett-Packard USA), 1 ANR project (France), 6 national projects (Spain), 6 regional projects (Galicia) and 1 European project.
Development of a performance model based on the Roofline Model for the FinisTerrae supercomputer. Department of Distributed and Dependable Systems. Charles University. Prague, Czech Republic.
Run-time code optimisation based on hardware counters. Department of Distributed and Dependable Systems. Charles University. Prague, Czech Republic.
Inspector/Executor adaptive strategies to efficiently parallelize N-Body simulation programs. Edinburgh Parallel Computer Centre (EPCC), The University of Edinburgh. United Kingdom.
Other Research Activities
Journal reviewer for Wiley Concurrency and Computation: Practice and Experience, the IET Information Security Journal, the Elsevier Future Generation Computer Systems journal and the International Conference on Parallel and Distributed Computing (EuroPar’14).
Co-organiser of the UNICO workshop (UsiNg and buIlding ClOud Testbeds) at the 5th IEEE International Conference on Cloud Computing Technology and Science (CloudCom 2013), Bristol.
Co-organiser of the BonFIRE infrastructure exhibition at the Future Internet Assembly Conference. (FIA 2013), Dublin.