Benchmarking TinyML Systems: Challenges and Direction
This work tackles the problem of benchmarking TinyML systems for researchers and developers, but it is incremental as it builds on existing efforts like TinyMLPerf.
The paper addresses the lack of a widely accepted benchmark for TinyML systems, presenting four benchmarks and discussing challenges and directions for developing a fair and useful hardware benchmark.
Recent advancements in ultra-low-power machine learning (TinyML) hardware promises to unlock an entirely new class of smart applications. However, continued progress is limited by the lack of a widely accepted benchmark for these systems. Benchmarking allows us to measure and thereby systematically compare, evaluate, and improve the performance of systems and is therefore fundamental to a field reaching maturity. In this position paper, we present the current landscape of TinyML and discuss the challenges and direction towards developing a fair and useful hardware benchmark for TinyML workloads. Furthermore, we present our four benchmarks and discuss our selection methodology. Our viewpoints reflect the collective thoughts of the TinyMLPerf working group that is comprised of over 30 organizations.