Hardware-aware Automated Architecture Search for Brain-inspired Hyperdimensional Computing

AeneasHDC: An Automatic Framework for Deploying Hyperdimensional Computing Models on FPGAs

“…On the algorithmic side, part of the design choices available for HDC models [17]- [19], [24] has been explored in [29] and [30]. The work in [29], presents HDTorch, a software library with CUDA extensions for testing the impact of different strategies on the model accuracy.…”

“…The work in [29], presents HDTorch, a software library with CUDA extensions for testing the impact of different strategies on the model accuracy. In HwAwHDC [30], a framework based on Reinforcement Learning optimization is used to search for the optimal trade-off between accuracy, model size, execution time, memory usage, and energy consumption of HDC software models implemented on embedded devices.…”