FastML Science Benchmarks: Accelerating Real-Time Scientific Edge Machine Learning

TL;DR

This paper introduces a set of benchmarks for ultra-fast machine learning at the scientific edge, aiming to guide hardware and software development for real-time data processing in scientific applications.

Contribution

It provides the first comprehensive benchmarks for real-time scientific edge ML, addressing the need for low-latency solutions amid hardware and data growth challenges.

Findings

Initial set of scientific ML benchmarks presented

Benchmarks cover diverse ML and embedded system techniques

Guidelines for designing future ultra-fast edge ML hardware

Abstract

Applications of machine learning (ML) are growing by the day for many unique and challenging scientific applications. However, a crucial challenge facing these applications is their need for ultra low-latency and on-detector ML capabilities. Given the slowdown in Moore's law and Dennard scaling, coupled with the rapid advances in scientific instrumentation that is resulting in growing data rates, there is a need for ultra-fast ML at the extreme edge. Fast ML at the edge is essential for reducing and filtering scientific data in real-time to accelerate science experimentation and enable more profound insights. To accelerate real-time scientific edge ML hardware and software solutions, we need well-constrained benchmark tasks with enough specifications to be generically applicable and accessible. These benchmarks can guide the design of future edge ML hardware for scientific applications capable of meeting the nanosecond and microsecond level latency requirements. To this end, we present an initial set of scientific ML benchmarks, covering a variety of ML and embedded system techniques.