hxtorch: PyTorch for BrainScaleS-2 -- Perceptrons on Analog Neuromorphic Hardware
This work provides a software framework for researchers and developers to accelerate neural network inference on neuromorphic hardware, but it is incremental as it builds on existing hardware and PyTorch interfaces.
The paper tackles the problem of using BrainScaleS-2 analog neuromorphic hardware as an inference accelerator for neural networks by integrating it into PyTorch, resulting in software that supports vector-matrix multiplications, convolutions, and hardware-in-the-loop training with automatic partitioning.
We present software facilitating the usage of the BrainScaleS-2 analog neuromorphic hardware system as an inference accelerator for artificial neural networks. The accelerator hardware is transparently integrated into the PyTorch machine learning framework using its extension interface. In particular, we provide accelerator support for vector-matrix multiplications and convolutions; corresponding software-based autograd functionality is provided for hardware-in-the-loop training. Automatic partitioning of neural networks onto one or multiple accelerator chips is supported. We analyze implementation runtime overhead during training as well as inference, provide measurements for existing setups and evaluate the results in terms of the accelerator hardware design limitations. As an application of the introduced framework, we present a model that classifies activities of daily living with smartphone sensor data.