I am currently pursuing my PhD in Computer Science at the University of California, Riverside. My research interests span in the areas of Parallel Computing, Parallel Architectures, Hardware Accelerated Architectures and Distributed & Large Scale Database Systems. I am member of the database laboratory at UCR and currently being advised by Professor Vassilis Tsotras. I am also working in collaboration with Professor Walid Najjar's group developing database applications on FPGA accelerated Platforms. Previously, i have been a member in the Smart Grid and FPGA/Parallel Computing Lab working under the supervision of Professor Viktor Prasanna at the University of Southern California. There, i participated in integrating power prediction models under a common API that would enable support for real time decission making during peak in energy demand. In my research, i used massively parallel architectures (i.e. GPUs) to support fast multi-dimensional and high-dimensional data processing.
I was hired to provide consultation in the development of parallel algorithms, mainly related to linear algebra and other numerical applications, on devices that utilize a large array of processors. My main responsibilities included benchmarking the available development tools and providing and abstract programming model for the new generation of accelerators.
I was the assigned TA of CS572 - Information Retrieval & Web Search Engines where i interacted with students through Piazza and held office hours to resolve any doubts related to the course subject. My responsibilities were to get acquainted with state of the art Large Scale Information Retrieval methodologies and systems that are currrently being used in popular web-search engines, in order to answer related questions from students.
During this period, i worked on problems related to high/multi-dimensional data processing using massively parallel architectures (i.e. GPUs). I focused on designing parallel algorithms that can achieve a balance between parallelism and work-efficiency. My research work was based on creating a parallel version of the support count operator that is widely used in Data mining.
I have been an active member in the Los Angeles Smart Grid Regional Demonstration Program researching technologies that address the real-time challenges associated with the ever growing energy demand in large urban areas such as the city of Los Angeles. This project has been a colloboration between the University of Southern California (USC), the Jet Propulsion Laboratory (JPL), the University of Southern California, Los Angeles (UCLA), the Los Angeles Department of Water & Power (LADWP) and the Department of Energy (DOE). I have been one of the developers of Demand Decission Support(DDS) API which was developed using Java to provide utilities real time decission making services.
Here i present certain side projects that i developed during my gradudate studies.
Developed entirely in Java, this is an extendable simulator that allows for developing and testing robot swarm algorithms. The example algorithms already implemented are random aggregation, collision detection using velocity obstacles. I used JOGL, JavaGeom libraries for my implementation. I was inspired by an implementation that i found here. A demonstration video is available here.