I am a computer science researcher whose interests lie primarily in the Systems domain. I have been a part of projects in Cloud Computing, Distributed Systems and Computer Architecture. My current research is focused on the following areas:
- Utilizing recent advances such as RDMA and Programmable Switches (P4) to accelerate large-scale system performance
- Analyzing the impact of Networking Failures on large-scale software systems
- Analysis of modern-day Serverless Computing infrastructure to identify performance bottlenecks and optimization opportunities
I am enrolled in the PhD program in Computer Science at the University of Waterloo as a part of the Waterloo Advanced Systems Lab (WASL). My supervisor is Dr. Samer Al-Kiswany. Prior to this, I was enrolled in the research-based master's program at Waterloo from 2019-2021 and I have been a Design Engineer at Advanced Micro Devices (AMD) from 2017-2019.
I received my master's degree in Computer Science from the University of Waterloo in 2021, working under the supervision of Dr. Samer Al-Kiswany. I received my undergraduate degree in Computer Science from PES University in Bangalore, India and have been supervised by Dr. K.V Subramaniam and Dr. Phalachandra. I was a research assistant at CCBD during my time there.
Complete Resume: Resume
LinkedIn: Sreeharsha Udayashankar
Partial Network Partitioning (ACM Transactions on Computer Systems 2022) [Paper]
- Basil Alkhatib, Sreeharsha Udayashankar, Sara Qunaibi, Ahmed Alquraan, Mohammed Alfatafta, Wael Al-Manasrah, Alex Depoutovitch, Samer Al-Kiswany
Orcbench: A Representative Serverless Benchmark (IEEE Cloud 2022) [Paper]
- Ryan Hancock, Sreeharsha Udayashankar, Ali Mashtizadeh and Samer Al-Kiswany
Benchmarking Diferentially Private Algorithms (TPDP - ICML 2021) [Paper]
- Huiyi Ning, Sreeharsha Udayashankar, Sara Qunaibi, Karl Knopf and Xi He
In-Network Scheduling for Real-Time Analytics (MMath Thesis - 2021) [Thesis]
- Sreeharsha Udayashankar
Falcon - Low Latency Network Accelerated Scheduling (EuroP4 - ACM CoNext 2020) [Paper]
- Ibrahim Kettaneh, Sreeharsha Udayashankar, Ashraf Abdel-Hadi, Robin Grosman and Samer Al-Kiswany
Gaming Modeling and Projections - The Impact of CPU Performance (AMD Asia Tech Conference 2019)
- Sreeharsha Udayashankar, Saumya Chandra and Don Cherepacha
In my spare time, I like to develop applications using a variety of languages and tools to keep my skills sharp. Some of these projects and their source code are showcased in this section.
MuZero on Super Mario Bros
I have attempted to evaluate MuZero, an algorithm for deep learning published by Google Deepmind in 2019, against Deep-Q-Networks with Double-Q-Learning, the state of the art model free algorithm. I have compared their performance using the game Super Mario Bros. The system architectural diagram has been shown below:
Clips of an agent trained using MuZero playing the game can be found below:
|Training GIF 1 (Level Complete)||Training GIF 2 (Level Complete)||Training GIF 3 (Level Fail)||Training GIF 4 (Level Fail)|
The Crypto Lottery
This project consists of an Ethereum-based back end and a front end powered by React.js. The objective of this project was to build a lottery application where users register using Ether and stand a chance to win the Jackpot.
The Jackpot is the prize pool consisting of contributions from all participants with the winner being chosen at regular time intervals. A participants chances also increase with the Ether they contribute up to a certain limit.
Python Game Collection
I have developed a few games in Python 3.6 using the PyGame library. Some of them are showcased below.
I enjoy running, cycling and hiking in my spare time. Here are a few images I've shot using my phone on some of my hikes:
|Sunshine Meadows - Banff National Park||Peyto Lake - Banff National Park|
|Flowerpot Island - Bruce Peninsula National Park||The Grotto - Bruce Peninsula National Park|
|Centennial Ridges - Algonquin Provincial Park||Track and Tower - Algonquin Provincial Park|
|BR Hills - Karnataka||Tadiandamol - Karnataka|