Collaborators

The research mentioned below is possible because of my wonderful collaborators:

• current students - {Vivek Shahare, Bonthu Vyuhita, Nithin Damarla}, former students - {R Eshwar, Rohit Garg, Soumyadeep Das, R Amogh, Parikshit Gehlaut, Kaushal Chavan, Abhishek Josyula, Pritesh Verma, Abhinay Anubola, Abhinav Verma}
• Dr. Milind Chabbi, my advisor Prof. Milind Kulkarni and the PLCL group at Purdue
• Prof. Amar Gaonkar, Prof. Amlan Barua. (IIT Dharwad)

Current Projects

• Parallelizing super-fast eigensolver for symmetric HSS matrices:ICPP'24 paper, Slides
• Parallel code generation for recursive linear algebra algorithms: ICPE'26 Paper, Slides
• Resiliency of locks: SPAA'23 Paper, Slides; ICS'26 Paper, Slides
• MLIR-based compiler for Python3.8
• Optimizing variadic functions

Past Projects

• D2P: From Recursive Formulations to Distributed Memory Codes. This work is about auto-generation of distributed-memory implementations of shared-memory algorithms. Paper (ACM DL)

(The technical report of a part of this project can be found here)
• SPIRIT - A framework for creating distributed recursive tree applications. The source code is available under the DM folder of Treelogy sourcecode. Paper (ACM DL)
• Treelogy - A Benchmark Suite and an Ontology for Tree Traversal Algorithms. Paper (IEEE Xplore)
• Hybrid CPU-GPU scheduling and execution of tree traversals - Scheduling optimization for tree traversal algorithms executing on GPUs. Paper (ACM DL)
• WaSP - Ensemble-based, Warm-Starting Parameter Initialization for Training of Deep Neural Network Models.