A memory hierarchical management system, method and computer readable storage medium

Abstract

In various information processing systems, it is often necessary to process both short-term real-time data and long-term historical data simultaneously. Traditional data storage solutions typically employ a single storage method. This approach suffers from drawbacks such as a conflict between storage capacity and access efficiency, a lack of semantic association, and interference from irrelevant data. While some solutions have attempted to use vector databases for semantic retrieval, they cannot simultaneously meet the requirements of real-time performance, completeness, and semantic association. In particular, existing database systems often assume deployment in the cloud or on server clusters; however, on resource-constrained local devices such as personal computers and smartphones, a lightweight, low-power, and privacy-secure memory management solution is needed. This application provides a memory hierarchical management system, method, and computer-readable storage medium, aiming to achieve efficient and scalable memory storage and retrieval while considering real-time response, long-term memory, and semantic association, and providing privacy protection and offline availability.

Images