Debashis Ganguly

Ph.D.

Department of Computer Science
School of Computing and Information
University of Pittsburgh
 Sennott Square, Office 5412
  210 S. Bouquet St.
  Pittsburgh, PA 15260
 (404) 834-5386
 debashis[at]cs{dot}pitt(dot)edu

I completed my PhD from Department of Computer Science, School of Computing and Information, University of Pittsburgh under the guidance of Dr. Rami Melhem (co-advised by Dr. Jun Yang). My dissertation was focused on designing adaptive memory management techniques for discrete CPU-GPU heterogeneous memory systems, including commodity cloud infrastructures and exascale systems. I have proposed a set of software-hardware co-operative solutions that leverage existing hardware advancements and software-managed unified memory runtime to relieve programmers from the burden of explicit memory management and to mitigate performance variability in consolidated heterogeneous systems.

My research interests are, broadly, computer architecture, heterogeneous systems, distributed computing, operating systems, and run-times. However, I am interested in studying many of the system-wide challenges, like QoS resiliency, fault-tolerance, encountered when building software for large-scale distributed systems.

I have started working as a Software and Hardware Modeling Engineer (ICT3) in Apple, Cupertino (IMG SW - GPU SW - Silicon Development Team) from November 2020.


Curriculum Vitae

Publications


Recent Publications

Conference Publications
DATE
[2021]
pdf ppt ppt Debashis Ganguly, Rami Melhem, Jun Yang, An Adaptive Framework for Oversubscription Management in CPU-GPU Unified Memory, 2021 Design, Automation & Test in Europe Conference & Exhibition. Feb 01-05, 2021. [TO APPEAR]
IPDPS
[2020]
pdf ppt ppt Debashis Ganguly, Ziyu Zhang, Jun Yang, Rami Melhem, Adaptive Page Migration for Irregular Data-intensive Applications under GPU Memory Oversubscription, The 34th IEEE International Parallel and Distributed Processing Symposium. May 18-22, 2020.
ISCA
[2019]
pdf ppt ppt Debashis Ganguly, Ziyu Zhang, Jun Yang, Rami Melhem, Interplay between Hardware Prefetcher and Page Eviction Policy in CPU-GPU Unified Virtual Memory, In Proceedings of the 46th International Symposium on Computer Architecture, pp. 224-235. 2019
CLOUD
[2017]
pdf ppt Debashis Ganguly, Mohammad H. Mofrad, Taieb Znati, Rami Melhem, John R. Lange, Harvesting Underutilized Resources to Improve Responsiveness and Tolerance to Crash and Silent Faults for Data-intensive Applications, In Cloud Computing (CLOUD) [Applications Track], 2017 IEEE 10th International Conference on (pp. 536-543). IEEE.
WACV
[2017]
pdf ppt Debashis Ganguly, Mohammad H. Mofrad, Adriana Kovashka, Detecting Sexually Provocative Images, In Applications of Computer Vision (WACV), 2017 IEEE Winter Conference on (pp. 660-668). IEEE.

Workshop Papers
ROSS
[2017]
pdf ppt Debashis Ganguly, John R. Lange, The Effect of Asymmetric Performance on Asynchronous Task Based Runtimes, In Proceedings of the 7th International Workshop on Runtime and Operating Systems for Supercomputers ROSS 2017 (p. 8). ACM.
WSSSPE
[2016]
pdf Debashis Ganguly, William Garrison, David Wilkinson, Bruce Childers, Adam Lee, Daniel Mosse, Demo: Composing, Reproducing, and Sharing Simulations, 4th Workshop on Sustainable Software for Science: Practice and Experiences (WSSSPE4).
ModSim
[2016]
ppt Debashis Ganguly, William Garrison, David Wilkinson, Bruce Childers, Adam Lee, Daniel Mosse, Composing, Reproducing, and Sharing Simulations, Workshop on Modeling & Simulation of Systems and Applications (ModSim).

Book Publication

Network and Application Security: Fundamentals and Practices; ISBN 978-1-57808-755-6; Science Publishers, Enfield, New Hampshire and CRC Press, Taylor & Francis Group

Undergraduate Publications

Passion