About Me:
I am a second-year PhD student in Electrical Engineering at University of Southern California (USC) co-advised by Shahin Nazarian and Paul Bogdan. My research interests lie in VLSI, Parallel Computer Architecture and Programming, Network on Chip, Complex Networks, and Machine Learning. My current research is on modeling applications as complex networks, mapping communities detected from complex networks onto NoCs, and designing high-performance NoCs.
Contact Information:
Research Projects
1. Complex Network Inspired Parallel Execution Optimization Framework
Applications can be described as weighted directed acyclic graphs. Nodes represent computations, and edges represent dependencies. Graph partitioning is an approach to divide the computations among processors, which remains vital in parallel computing. We are inspired from the complex network theory where there is an increasing need to find community structures, clusters of densely connected network nodes with sparse inter-community edges by an optimization model.

The paper proposes a new optimization framework to partition graphs into communities with load balancing and minimal communication overhead for exascale multicore systems.

This paper decides the best data placement strategy in Processing-in-Memory (PIM) via multilayer networks to exploit the high throughput.

Publications (Google Scholar)
Available Tools
Address: 3740 McClintock Ave, Los Angeles, CA 90089