Extreme data rate scientific experiments create massive amounts of data that require efficient ML edge processing. This leads to unique validation challenges for VLSI implementations of ML algorithms: enabling bit-accurate functional simulations for performance validation in experimental software frameworks, verifying those ML models are robust under extreme quantization and pruning, and enabling ultra-fine-grained model inspection for efficient fault tolerance. We discuss approaches to developing and validating reliable algorithms at the scientific edge under such strict latency, resource, power, and area requirements in extreme experimental environments. We study metrics for developing robust algorithms, present preliminary results and mitigation strategies, and conclude with an outlook of these and future directions of research towards the longer-term goal of developing autonomous scientific experimentation methods for accelerated scientific discovery.
@article{arxiv.2406.19522,
title = {Reliable edge machine learning hardware for scientific applications},
author = {Tommaso Baldi and Javier Campos and Ben Hawks and Jennifer Ngadiuba and Nhan Tran and Daniel Diaz and Javier Duarte and Ryan Kastner and Andres Meza and Melissa Quinnan and Olivia Weng and Caleb Geniesse and Amir Gholami and Michael W. Mahoney and Vladimir Loncar and Philip Harris and Joshua Agar and Shuyu Qin},
journal= {arXiv preprint arXiv:2406.19522},
year = {2024}
}