ProbLP: A framework for low-precision probabilistic inference
This addresses energy efficiency for edge devices performing probabilistic inference, but it is incremental as it builds on existing low-precision and hardware design concepts.
The paper tackles the problem of energy efficiency in Bayesian probabilistic inference for edge devices by proposing ProbLP, a framework that automates the design of low-precision hardware, resulting in validated improvements on embedded-sensing benchmarks.
Bayesian reasoning is a powerful mechanism for probabilistic inference in smart edge-devices. During such inferences, a low-precision arithmetic representation can enable improved energy efficiency. However, its impact on inference accuracy is not yet understood. Furthermore, general-purpose hardware does not natively support low-precision representation. To address this, we propose ProbLP, a framework that automates the analysis and design of low-precision probabilistic inference hardware. It automatically chooses an appropriate energy-efficient representation based on worst-case error-bounds and hardware energy-models. It generates custom hardware for the resulting inference network exploiting parallelism, pipelining and low-precision operation. The framework is validated on several embedded-sensing benchmarks.