Avatar

Marcelo Orenes Vera

Computer Science PhD Candidate

Princeton University

Biography

I am a first-year graduate student in Computer Science at Princeton University advised by Professor Margaret Martonosi. My broad area of research is Computer Architecture and I am a contributor to the DECADES, a full-stack system design that aims to improve the performance, power, programmability, and portability of several emerging workflows in the broad areas of Machine Learning and Graph Analytics. Additionally, this platform rapidly adapts to the increasingly flexible and blurred boundary between software and hardware through various reconfigurable hardware features, depending on application characteristics. Currently, I am exploring the hardware design part of the project in conjunction with Professor David Wentzlaff group, and the software part, specially the simulator and the programming model, with Margaret Martonosi group.

Before joining the Graduate School I worked at Arm in its office in Norway, where I contibuted to the development of the Mali GPU. I worked there as a Hardware Engineer taking responsability on Formal Verification and Hardware Design inside the Texture Mapper Unit. Prior working at Arm I mainly focused on Software. I worked with Professor Mercedes Valdes Vela at Murcia University, where I received my bachelors degree in Computer Science. My research involved creating an indoor location system based on Bluetooth 4.0 LE technology to guide users in museums with the help of a mobile App, to after analyse the patterns of user trajectories inside the building. We leverage techniques like Clustering to profile users based on their movement behaviour and App interaction. We personalized the user experience based on the profile detected by the ML algorithm.

I enjoy programming and have served as a software engineering, developing mobile apps in iOS and Android platforms at SecondLemon and OBT Control, in Murcia, Spain. ch was focused on the development of computer simulations to model real-world observations of these cosmic events.

Outside of research, I enjoy playing soccer, doing rock climbing, cycling and doing cross-country ski during the winter.

Interests

  • Computer Architecture
  • Heterogeneous parallelism
  • Software-defined hardware and domain-specific architectures
  • Hardware Design

Education

  • PhD in Computer Architecture, Current

    Princeton University

  • BSc in Computer Science, 2017

    University of Murcia (Spain)

  • Exchange year Computer Science, 2016

    University of Hasselt (Belgium)

Projects

DECADES

A DARPA-funded project to create specialized, reconfigurable hardware to accelerate important applications.

Experience

 
 
 
 
 

PhD candidate

Princeton University

Sep 2019 – Present Princeton, New Jersey
  • Designing full-stack approaches to optimize the performance, power, and programmability of graph and other sparse analytics that are the heart of many modern big data applications; contributing to the development of a specialized, reconfigurable hardware platform for accelerating different software applications as part of DARPA’s Software Defined Hardware (SDH) program
 
 
 
 
 

Hardware engineer

Arm Ltd.

Jul 2017 – Aug 2019 Trondheim, Norway
  • Investigation and design of optimizations to the current implementation of several modules within the Texture Mapper of the Mali GPU, specifically the Texture Cache and the miss-path of it. Multilevel cache systems development and replacement policies optimization for performance and bandwidth saving. I also worked in the main pipeline of data access in an out of order execution architecture.
  • Formal Verification at Unit Level and Bug hunting approaches for liveness properties between units. UVM usage for simulation testbench of the texturing unit.
 
 
 
 
 

Undergraduate Research Assistant

University of Murcia

Feb 2016 – Jul 2017 Murcia, Spain
  • Developing a platform for indoor location based on Bluetooth 4.0 LE technology and analysis of user trajectories making use of clustering techniques.

Contact