Rising Stars 2020:

Rasha Karakchi

Postdoctoral Researcher

University of South Carolina


PhD '20 University of South Carolina

Areas of Interest

  • Computer Architecture and Engineering

Poster

An Overlay Architecture for Pattern Matching

Abstract

Deterministic and Non-deterministic Finite Automata (DFA and NFA) comprise the fundamental unit of work for many emerging big data applications, motivating recent efforts to develop Domain-Specific Architectures (DSAs) to exploit fine-grain parallelism available in automata workloads. Our research involves conceptual and applied components. The applied part involves design a reusable customized automata processor overlay on a Field-Programmable Gate Array (FPGA)- based platform. This design can maximally exploit on-chip memory parallelism for NFA evaluation. The conceptual part is the application of novel hardware allocation algorithms to a complex space of design trade-offs in both the computational and communication aspects of the design. Our research continues to adapt the practical components of Automata Processor design to operate in the cloud, in addition to develop techniques to improve the programmable routing of the design and reduce the output latency.

Bio

Rasha Elham Karakchi is a postdoctoral researcher at University of South Carolina, working with Dr. Jason Bakos. Her research interests include domain-specific architectures, reconfigurability, computer architecture and systems on chip. Rasha holds a Ph.D. degree in computer science and computer engineering and a M.E. degree in computer engineering from University of South Carolina. Before that, Rasha obtained her M.Sc. degree and B.Sc. degree in computer engineering from Mosul University-Iraq. She is a recipient of SPARC Graduate Research Grant from University of South Carolina from 2016-2017.