Cortical Processing with Thermodynamic-RAM
This work presents an incremental hardware advancement for machine learning applications, offering potential power and speed benefits over traditional von Neumann architectures.
The authors developed Thermodynamic-RAM (kT-RAM), a memristor-based neural processing unit that combines memory and processing, and demonstrated its application in machine learning benchmarks, achieving a 99.5% F1 score on an MNIST classification task with an improved training method.
AHaH computing forms a theoretical framework from which a biologically-inspired type of computing architecture can be built where, unlike von Neumann systems, memory and processor are physically combined. In this paper we report on an incremental step beyond the theoretical framework of AHaH computing toward the development of a memristor-based physical neural processing unit (NPU), which we call Thermodynamic-RAM (kT-RAM). While the power consumption and speed dominance of such an NPU over von Neumann architectures for machine learning applications is well appreciated, Thermodynamic-RAM offers several advantages over other hardware approaches to adaptation and learning. Benefits include general-purpose use, a simple yet flexible instruction set and easy integration into existing digital platforms. We present a high level design of kT-RAM and a formal definition of its instruction set. We report the completion of a kT-RAM emulator and the successful port of all previous machine learning benchmark applications including unsupervised clustering, supervised and unsupervised classification, complex signal prediction, unsupervised robotic actuation and combinatorial optimization. Lastly, we extend a previous MNIST hand written digits benchmark application, to show that an extra step of reading the synaptic states of AHaH nodes during the train phase (healing) alone results in plasticity that improves the classifier's performance, bumping our best F1 score up to 99.5%.