Posit arithmetic in 8-16 bits achieves comparable accuracy to 16-32 bit floats in cough detection and ECG R-peak tasks, with PHEE RISC-V hardware showing 38% smaller area and 42.3% less power in 16nm synthesis.
BiomedBench: A bench- mark suite of TinyML biomedical applications for low-power wear- ables,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Increasing the Energy-Efficiency of Wearables Using Low-Precision Posit Arithmetic with PHEE
Posit arithmetic in 8-16 bits achieves comparable accuracy to 16-32 bit floats in cough detection and ECG R-peak tasks, with PHEE RISC-V hardware showing 38% smaller area and 42.3% less power in 16nm synthesis.