The methodology demonstrates energy-efficient hardware-software co-design in multi-core systems, suggesting improved neural network performance.
The possibilities of intelligent multi-core Cyber-Physical System on Chip (CPSoN) for providing high-quality properties of universality, scalability and energy-efficiency based on modern VLSI K1879ВМ8Я and software Neuro Matrix are shown. A methodology for intelligent management of neuroprocessor resources is proposed for the purpose of joint design of hardware and software (Co-Design) based on Intellectual Multiprocessing and set-theoretical equivalence of a group of algorithms, a group of cores and a group of robots on a chip. Mathematical formalization of the analysis of various structures is carried out for the purpose of an optimal choice allowing us to design software-reconfigurable multi-core neuroprocessor architecture for processing large data flows. Modeling multi-criteria control for choosing the best structure according to various technical characteristics is carried out with emphasis on significant speed up the work of neural network tools, accelerators and servers.
No takes yet. Share an insight, caveat, or question.
Ruchkin et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: