Sriram Devata

I am a CS PhD student at UIUC, and am interested in developing hardware and software for datacenters that are robust to changes for emerging software paradigms.

Extended Reality (XR - encompassing AR/VR/MR) is one such paradigm that is transforming how we interact with computing. Current XR devices have a long way to go for their widespread adoption and sustained full-day usage. When considering offloading XR tasks to a datacenter, these tasks demand low latency and have strict performance standards to provide a high-quality user experience. I work with Prof. Sarita Adve to develop the hardware architecture and/or software systems for the cloud to facilitate the offloading of XR workloads.

Before my PhD, I worked with Prof. Akshitha Sriraman and Dr. Gilles Pokam to improve datacenter efficiency by characterizing datacenter tax operations. This work allowed us to recognize common low-level software primitives across datacenter tax operations, and also helped identify combined acceleration opportunities.

Previously, I also worked on using Machine Learning to solve problems in Computational Chemistry at IIIT-H and University of Waterloo.