StreamBrain: An HPC Framework for Brain-like Neural Networks on CPUs, GPUs and FPGAs
This work enables practical use of BCPNN, a less-explored machine learning algorithm, for researchers and practitioners in high-performance computing, though it is incremental as it adapts an existing method to new hardware.
The authors tackled the challenge of deploying brain-like Bayesian Confidence Propagation Neural Networks (BCPNN) in high-performance computing systems by introducing StreamBrain, a domain-specific language framework that supports CPUs, GPUs, and FPGAs, achieving training on MNIST within seconds and demonstrating BCPNN on STL-10 size networks for the first time.
The modern deep learning method based on backpropagation has surged in popularity and has been used in multiple domains and application areas. At the same time, there are other -- less-known -- machine learning algorithms with a mature and solid theoretical foundation whose performance remains unexplored. One such example is the brain-like Bayesian Confidence Propagation Neural Network (BCPNN). In this paper, we introduce StreamBrain -- a framework that allows neural networks based on BCPNN to be practically deployed in High-Performance Computing systems. StreamBrain is a domain-specific language (DSL), similar in concept to existing machine learning (ML) frameworks, and supports backends for CPUs, GPUs, and even FPGAs. We empirically demonstrate that StreamBrain can train the well-known ML benchmark dataset MNIST within seconds, and we are the first to demonstrate BCPNN on STL-10 size networks. We also show how StreamBrain can be used to train with custom floating-point formats and illustrate the impact of using different bfloat variations on BCPNN using FPGAs.