Keep it local: architectures and models of computation for embedded data processing near the source

December 11, 2020 at 11:00 AM to 12:00 PM

Data Science Distinguished Speaker Seminar Series  

Abstract

The advent of big data has created new constraints on communication and analytics layers. The staggering amounts of data cannot be communicated in their raw form without overburdening communication bandwidth, and cannot be processed in a central location without overburdening processing capabilities. Thus, there has been a push towards more distributed embedded processing, such that data can be transformed into information near the source, minimizing communication and distributing the computation effort.

In this talk, we’ll examine architectures and models of computation for embedded systems responsible for near-source data processing (i.e., the edge in edge-computing) and examine recent advances for optimizing the performance and power consumption of these systems, including heterogeneous multi-core processors, hardware acceleration, new models of parallel computation and algorithmic refactoring.

About the Speaker

Dr. Garcia joined Carleton University in 2018, establishing a research program on embedded systems. Prior to joining Carleton, Dr. Garcia was a Senior Research Scientist at the United Technologies Research Center (Cork, Ireland), working on safety-critical embedded systems and a Post-Doctoral Research Associate at Heriot Watt University (Edinburgh, U.K.), working on embedded sensors for signal processing. Dr. Garcia was awarded his PhD by the University of Minho, Portugal, in 2015; during his doctoral work, Dr. Garcia was a visiting researcher at the University of Wurzburg, Germany, and the Asian Institute of Technology, Thailand.

Dr. Garcia’s research interests include embedded systems design and programming, models of computation, computer architectures and reconfigurable computing. He is a member of the IEEE and a regular reviewer for several journals and conferences.

Zoom webinar information for the Data Science Distinguished Speaker Seminar Series is sent to our mailing list (sign up here). If you are not on our mailing list and would like to attend this virtual seminar, please e-mail cuids@carleton.ca.