Tomé Maseda Dorado
- PhD Student & Teaching assistant
- A Coruña, Galicia, Spain
About me
Hi! I am Tomé Maseda Dorado. I hold a Bachelor’s degree in Computer Engineering and a Master’s degree in High Performance Computing, both from the University of A Coruña (UDC). I am currently pursuing a PhD in Computer Science as a member of the Computer Architecture Group (GAC), with my research focusing on the study, implementation, and evaluation of techniques for dynamic energy management in container-based environments. In parallel, I work as a teaching assistant at UDC, where I teach Computer Architecture .
Throughout my academic path, I have specialised in virtualisation technologies such as Docker and Apptainer/Singularity , as well as Infrastructure as Code (IaC) tools like Vagrant and Ansible . I also have strong experience in designing real-time systems using Bash and Python , combined with data analysis using libraries such as NumPy , Pandas , and Scikit-learn .
In addition to my main areas of expertise, I have worked with Big Data frameworks including Hadoop , Spark , and Flink . I also have experience in the design and optimization of parallel and distributed algorithms across both HPC and Big Data environments. Furthermore, I have explored several Machine Learning techniques, including linear and polynomial regression, decision trees, and neural networks.
My main research interests are closely aligned with these areas and include serverless computing, virtualised environments, power consumption management, and data science applied to real-time systems. I also maintain a broader interest in algorithm optimisation and parallelisation, information security, and Big Data.
Research Projects
CPUPowerWatcher is a framework designed to automate CPU monitoring and time-series data collection while executing workloads. Its main purpose is to generate high-quality datasets for the training and evaluation of power consumption models.
CPUPowerWatcher automatically deploys monitoring daemons (e.g., PAPI, cpumetrics) together with an InfluxDB time-series database. It then runs user-defined workloads, which can range from synthetic stress tests (using tools like stress-ng) for the training phase, to well-known High Performance Computing benchmarks (e.g., NPB, Fio), or even real-world applications (e.g., Smusket) for testing. This framework enables fine-grained control over synthetic workloads by exposing parameters such as the stress pattern, which defines how CPU utilisation evolves over time during the experiment (e.g., stairs-up, zigzag, uniform), and the core distribution, which determines the specific order in which CPU cores are allocated to workloads when its usage requirements changes (e.g., Group_P&L, Spread_P&L).
Work Experience
University of A Coruña
Teaching assistantTeaching assistant of Computer Structure at University of A Coruña .
University of A Coruña
Predoctoral researcherResearcher at the Computer Architecture Group during the development of my PhD thesis.
INRIA (Lille, France)
Visiting scientistResearch stay during my PhD to collaborate with the Spirals team .
Research center on ICT (CITIC)
Research assistantDesign, implementation and optimization of parallel and distributed algorithms in High Performance Computing and Big Data environments.
Minsait
Bootcamp AnalyticsInternship contract. Basic management of common enterprise data storage architectures (e.g. data warehouses, data marts) and ETL tools. I specialised in data visualisation using Microsoft Power BI.
Education
University of A Coruña
Doctor of Philosofy (Ph.D.) in Computer SciencePhD Thesis: Techniques for dynamic energy management as a resource in container-based environments.
University of A Coruña
Master's degree in High Performance ComputingMasther’s Thesis: Tool for modeling processor power consumption using software
Masther’s Degree Academic Award. Average score: 9.73
University of A Coruña
Bachelor's degree in Computer EngineeringBachelor’s Thesis: Implementation and evaluation of TPCx-HS benchmark using Big Data processing technologies