An AI intelligent computing platform reasoning resource scheduling method and system

By collecting memory allocation records and communication quality indicators to construct a weighted bipartite graph, the resource scheduling of the AI ​​computing platform is optimized, solving the problems of memory fragmentation and insufficient communication quality awareness, improving task execution efficiency and system throughput, and ensuring stability and response speed in high-concurrency scenarios.

CN122111667APending Publication Date: 2026-05-29BEIJING YIYONG TIMES TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
BEIJING YIYONG TIMES TECH CO LTD
Filing Date
2026-02-24
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Existing AI computing platforms lack awareness of memory fragmentation and communication quality in large-scale clusters, resulting in low execution efficiency of inference tasks and reduced system throughput, especially affecting stability and response speed in high-concurrency scenarios.

Method used

By acquiring task load information, memory allocation records, and retransmission event time sequences from the bus data link layer, Shannon entropy is used to calculate memory fragmentation and communication congestion index. A weighted bipartite graph is constructed, and the Hungarian algorithm is used to optimize resource scheduling, avoiding the allocation of tasks to unstable nodes and achieving a balance between memory continuity and communication stability.

Benefits of technology

It improves the throughput and stability of the AI ​​computing platform under high concurrency load, reduces task response latency, avoids memory fragmentation and communication interference, and ensures optimized resource scheduling and efficient execution.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122111667A_ABST
    Figure CN122111667A_ABST
Patent Text Reader

Abstract

The application provides an AI intelligent calculation platform reasoning resource scheduling method and system, relates to the technical field of computer resource management and task scheduling, and obtains task load information of a to-be-reasoned task, memory allocation records of an AI intelligent calculation platform in a preset time window, and a retransmission event timing sequence of a bus data link layer; obtains the discrete entropy of each computing node through a Shannon entropy formula according to the memory allocation records; counts the retransmission times of the retransmission event timing sequence in a sliding time window to generate a communication congestion index of each computing node; determines a weight coefficient according to the task load information, constructs a weighted bipartite graph, and determines a matching edge set that satisfies the most matching edges and the smallest sum of edge weights in the weighted bipartite graph by using a Hungarian algorithm, so as to be used for resource scheduling of the to-be-reasoned task. The problem of throughput decline caused by serious memory fragmentation and communication interference in the large model concurrent reasoning scene is solved.
Need to check novelty before this filing date? Find Prior Art