I am a final-year PhD student in the UBC SoC Research group at the University of British Columbia advised by Steve Wilton. Having previously worked in the FPGA domain, my current research interests lie in improving reliability/trust in tinyML systems. In addition, I work with my research group on improving CAD algorithms for FPGAs with ML techniques. Also, I’m part of the Climate Change AI community, specifically interested in harnessing the power of IoT/tinyML for efficient carbon capture.

Previously, I spent two amazing years at Microsoft Research India advised by Vivek Seshadri and Rahul Sharma. Before that, I earned my Masters at University of Florida. During my time as a grad student and as a Research Engineer, I had the privilege to collaborate with CERN on the Large Hadron Collider’s CMS experiment under the supervision of Herman Lam and Darin Acosta.

Chekout my CV for more details.

Publications

[MlforSys'24] DEBUG-HD: Debugging TinyML models on-device using Hyper-Dimensional computing
Nikhil P Ghanathe, Steven J E Wilton
Machine Learning for Systems Workshop at NeurIPS , 2024

[Under Review] QUTE: Quantifying Uncertainty in TinyML with Early-exit-assisted ensembles for model monitoring
Nikhil P Ghanathe, Steven J E Wilton
Preprint , 2024

[MLCAD '24] Enabling Risk Management of Machine Learning Predictions for FPGA Routability
Andrew David Gunter, Maya Thomas, Nikhil P Ghanathe, Steven J E Wilton
2024 ACM/IEEE International Symposium on Machine Learning for CAD (MLCAD), 2024

[CF '23] T-RecX: Tiny-Resource Efficient Convolutional neural networks with early-eXit
Nikhil P Ghanathe, Steve Wilton
International Conference on Computing Frontiers (CF), 2023

[FPL '21] MAFIA: Machine Learning Acceleration on FPGAs for IoT Applications
Nikhil P Ghanathe, Vivek Seshadri, Rahul Sharma, Steve Wilton, Aayan Kumar
International Conference on Field-Programmable Logic and Applications (FPL), 2021

[PLDI '19] Compiling KB-sized machine learning models to tiny IoT devices
Sridhar Gopinath, Nikhil P Ghanathe, Vivek Seshadri, Rahul Sharma
ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI), 2019

[TWEPP '16] Software and firmware co-development using high-level synthesis
Nikhil P Ghanathe, Alexander Madorsky, Herman Lam, Darin Acosta et al. on behalf of CMS collaboration
Topical Workshop on Electronics for Particle Physics (TWEPP), 2016