Abstract

We present a unified integer-only kernel for hyperbolic functions tanh(x) and sigmoid(x), designed for embedded RISC-V platforms without floating-point units. The kernel combines LUT anchoring, CORDIC microrotations, and linear correction in Q20 arithmetic, producing outputs directly in Q1.16 format suitable for ML inference. Exhaustive evaluation across the entire 17-bit input space demonstrates deterministic accuracy better than 1 ULP in Q1.16, with maximum absolute errors below . Compared to software-emulated floating-point implementations, our approach achieves significant speedup while reducing hardware complexity by unifying multiple activation functions in a single IP block. This makes the method highly relevant for efficient deployment of neural networks on resource-constrained RISC-V microcontrollers.

Date

2026

Authors

Kamil Kaczmarski, Pawel Gepner, Ewa Deelman, Pawel Poczekajlo, Leonid Moroz, Nataliia Gavkalova

Book

International Conference on Computational Science

Pages

573-587

Publisher

Springer Nature Switzerland