tinyMLPerf: Benchmarking Ultra-low Power Machine Learning Systems Post date February 12, 2021 ← Optimizing Inference Efficiency for Tiny DNNs → Using ML for ML to Span the Gamut of TinyML Hardware