About
I am Devansh Garg, an M.Tech student in the Department of Computer Science and Engineering at the Indian Institute of Technology Hyderabad (IIT Hyderabad), advised by Dr. Praveen Tammana. Before joining IIT Hyderabad I received my B.Tech in Computer Science from NIT Goa in 2025.
My research is in systems for large-scale machine learning: improving the efficiency of training and inference. I currently work on LLM inference optimization and in-network computing, alongside datacenter congestion control and programmable data planes. Most of my work is evaluated on real testbeds, including BMv2 and Intel Tofino. A complete list of publications is available on my Publications page and Google Scholar.
I am looking for summer 2027 research internships in ML systems and networking.
Education
- M.Tech in Computer Science and Engineering, IIT Hyderabad, Jul. 2025 – May 2028 (expected)
- B.Tech in Computer Science and Engineering, NIT Goa, Nov. 2021 – May 2025
Research Interests
- Systems for AI/ML — training and inference efficiency
- In-network computing
- Datacenter networking, congestion control, and RDMA
- Programmable data planes
Selected Publications
IFIP NET’25Detecting Manipulation to Table Rules in the Programmable Data Planes
IFIP Networking Conference (IFIP Networking), 2025.
OCIT’25XF-SISS: XOR-Based Fault Tolerant and Flexible Secret Image Sharing Scheme
OITS International Conference on Information Technology (OCIT), IEEE, 2025.