Suman Ghosh
Graduate Research Assistant at Georgia Tech School of Electrical & Computer Engineering- Claim this Profile
Click to upgrade to our gold package
for the full feature experience.
Topline Score
Bio
Experience
-
Georgia Tech School of Electrical & Computer Engineering
-
United States
-
Higher Education
-
1 - 100 Employee
-
Graduate Research Assistant
-
Jan 2023 - Present
Working under the guidance of Dr. Linda S Milor on Detection and Prevention of Aging Attacks Targeting Electromigration, sponsored by Global Research Collaboration (GRC) | Hardware Security | ICT Security Challenges. Current goals are to study and model the effects of electromigration in interconnects to get insights and possibly predict the MTTF for given interconnect geometries from experimental data provided. To summarize, my primary focus is on developing models and conducting simulations to analyze the hydrostatic stress in interconnect geometries. Additionally, I am working on creating MARS models to calibrate the system. Show less
-
-
-
SRC Research Scholars Program
-
United States
-
Research Services
-
200 - 300 Employee
-
Research Scholar
-
Jan 2023 - Present
Detection and Prevention of Aging Attacks Targeting Electromigration Global Research Collaboration (GRC) | Hardware Security | ICT Security Challenges With the down-scaling of process nodes, we have started to see intentional wearout and aging attacks. Such attacks are especially dangerous when they are launched remotely through software by malicious users. They can be as sever as Row Hammer attacks which target memory elements containing security sensitive information (security policies, security keys, passwords, etc.). One potential source of aging attacks is electromigration. Electromigration is an increasing concern for advanced process nodes as current densities in interconnect increase with scaling. Like Row Hammer, wearout due to electromigration, if properly targeted, can affect a microprocessor's ability to preperly manage sensitive information. This project builds on our work in simulating the electromigration degradation process in microprocessors running complex workloads (such as simulating the ARMv8-A architecture running benchmarks with gem5) and on work by others who have found methods to access sensitive information through malicious software that is launched remotely. Our electromigration simulation flow is based on physical models, including the state-of-the-art hydrostatic stress simulation methodology to identify voiding locations, and microprocessor emulation to determine current profiles during operation. The simulation flow involves machine learning, which must be expanded to characterize electromigration in the irregular geometries in all long interconnects in a microprocessor. My current focus centers on the development of modelling and simulations of hydrostatic stress in interconnect geometries, in addition to the creation of MARS models for the calibration system. Show less
-
-
-
Cognizant
-
Uzbekistan
-
Education
-
1 - 100 Employee
-
Data Engineering Intern
-
Mar 2022 - Jun 2022
Service Line: Data | AWS • Data analytics: Python, SQL, and Spark courses. Completed Data Engineering course from Databricks. • Got assigned to AWS track for Data service-line. Used EC2, S3, VPC, Lambda, Redshift, and other AWS services. Learnt from AWS experts how to implement different services for business analytics for clients from AWS in live projects for Cognizant during hands-on sessions. • Mini project using VPC and RDS service of AWS to setup private access Database and Database server for a model company. Show less
-
-
Education
-
Georgia Institute of Technology
Master's degree, Electrical and Computer Engineering -
Vellore Institute of Technology
B.Tech, Electrical and Electronics Engineering -
South Point High School, Kolkata
Senior School Certificate Examination, Physics Chemistry Mathematics