About Me

Welcome to my personal website! My name is Sergi Siso, I hold a PhD in Computer Science and I am currently working as a High-Performance Software Engineer at the Hartree Centre, which is part of the Science and Technology Facilities Council (STFC), located in Daresbury Laboratory, UK. In this webpage you will find my curriculum as well as a list of my publications and open-source software contributions. You can also find a PDF version of my curriculum here.


Research Interests

My research interests are in the field of High-Performance Computing (HPC), particularly in the portability and optimisation of scientific and engineering applications for modern highly-parallel architectures. My research focuses on leveraging compilation technologies such as Domain Specific Languages, Source-to-Source translators, and Runtime Compilation to develop software solutions that provide optimal performance while maintaining good productivity and maintainability of the application domain characteristics. I am one of the main authors of PSyclone, a source-to-source optimising Fortran compiler designed to parallelise and accelerate weather and climate applications, such as the next generation atmospheric modelling system utilised by the UK MetOffice.


Education

  • 2016 - 2023: Ph.D. in Computer Science, University of Liverpool

    • Thesis: "Augmenting Compiler Optimisations through Source-to-Source Injection of Runtime Information-aware Transformations"
    • Supervisors: Dr. Jeyan Thiyagalingam and Dr. Frans Coenen
  • 2013-2014: MSc with Distinction in High Performance Computing, EPCC, University of Edinburgh

    • Scholarship: "Highly Skilled Workforce" award of the Scottish Funding Council.
    • Dissertation: "Parallelisation of the Coupled Coherent States quantum dynamics simulation"
    • Advisor: Dr. Andrew Turner
  • 2011-2013: Master's Degree in Information Technology, Polytechnic University of Catalonia

    • Dissertation: "Simulating parallel systems using summarized application information"
    • Advisor: Dr. Jesus Labarta
  • 2008-2011: Bachelor Degree with Honours in Enginyeria Tècnica en Informàtica de Sistemes (Computer engineering), University of LLeida

    • Dissertation: "Electronic voting system over elliptic curve cryptography"
    • Advisor: Dr. Josep M. Miret


Experience

  • Feb 2015 - present: High Performance Software Engineer at Hartree Centre, Science and Technology Facilities Council (UKRI), Daresbury laboratory.

    • I lead the Intel Parallel Computer Centre (IPCC) at Hartree Centre. This involved modernising several engineering applications used by UK research and industry communities to perform optimally on the Intel manycore processors. In addition, I was a member of Intel eXtreme Performance User Group (IXPUG) steering committee.
    • I contributed to the performance analysis and optimisation of multiple applications such as DualSPHysics (a smooth particle hydrodynamics to study free-surface flow phenomena), DL_MESO (a mesoscale Lattice Boltzmann mesoscopic simulation package), AU3D (an industrial FE solver to study unsteady compressible flow and structural dynamics used by Rolls-Royce plc), and participated in international research projects such as EuroEXA (a EU Horizon 2020 effort to co-desing ExaScale applications)
    • Currently I am a contributor to the PSyclone source-to-source optimising Fortran compiler used to parallelise codes used by the weather and climate community, such as LFRic (an Earth Weather and Climate modelling system) and NEMO (an ocean modelling system).
  • Feb 2013 - Aug 2013: Resident Student Researcher at the Performance Tools Department of Barcelona Supercomputing Centre

    • I developed software to process summarized profiling information from highly scalable applications and format such information in a way that the Dimemas software is able to simulate its performance in different systems. It was part of the CEPBA-tools toolchain.
  • 2012 – 2013: Developer at SERIMAG MEDIA SL

    • I developed a document classifier for a banking company. The application used data mining technologies such as Support Vector Machines and the Viterbi Algorithm to classify a stream of incoming documents automatically.
  • Dec 2010 - Aug 2011: Grant-holder as a cryptography researcher at the Department of Cryptography and Graphs, University of LLeida

    • I developed an electronic voting system which works on different types of homomorphic cryptosystems, specifically over elliptic curves. Studied and implemented other cryptography topics.