I am a third year Ph.D. student at the Systopia Lab at University of British Columbia working with Alexandra Fedorova and Margo Seltzer. My research interests lie in Operating systems, Architecture and Systems performance.
I completed my Masters in Computer Science at University of Massachusetts, Amherst. During my time there I worked at PLASMA with Emery Berger and Hui Guan.
Prior to UMass I worked at Morgan Stanley on improving performance of distributed applications.
I completed my undergraduate from Nirma University in Ahmedabad.
Operating systems, Architecture, Systems performance.
Publications
CHERI-picking: Leveraging capability hardware for prefetching. [2023] [PDF] Shaurya Patel , Sid Agrawal, Alexandra (Sasha) Fedorova, Margo Seltzer
Programming Languages and Operating systems workshop (PLOS'23) colocated with SOSP 2023.
FreeLunch: Compression-based GPU Memory Management for Convolutional Neural Networks. [2021] [PDF] Shaurya Patel , Tongping Liu, Hui Guan
IEEE/ACM Workshop on Memory Centric High Performance Computing (MCHPC)
Research
My research interests lie in the intersection of Operating systems and hardware. Specifically, I am interested in exploring how better abstractions can be created in Operating systems that can leverage new hardware to improve systems performance.
During my Masters I was a part of the PLASMA group and was
advised by Emery Berger. My project explored how existing thread scheduling algorithms can be improved by using statistical techniques.
I also worked with Hui Guan at UMass. As a part of this project I proposed a new compression based memory management policy for ML training.
• Updated a cache library for indexing location data being used by multiple teams to reduce latency by upto 6% across multiple projects.
• Added a feature to low-latency trades serving infrastructure at google to enable faster auctions of ads to be served by early rejection of 5% of total ads.
• Optimized the scalability distributed trading platform. Used containerization, machine learning and improved load balancing. Successfully increased throughput by 300%.
• Developed a distributed platform to calculate the fees for a given trade being processed using Gemfire. This increased efficiency of business processes by 200%.
Awards and Grants
OSDI and PLMW Student Grant, 2020
Morgan Stanley Global Tech Excellence award, 2018.
Morgan Stanley Outstanding delivery award, 2018.
Morgan Stanley Best Project award, 2018.
Regional mathematical Olympiad scholarship (3rd in city), 2009.