An adaptive partition decision task scheduling method and system based on an integrated computing architecture

CN121918971BActive Publication Date: 2026-06-09QILU UNIVERSITY OF TECHNOLOGY (SHANDONG ACADEMY OF SCIENCES) +1
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
QILU UNIVERSITY OF TECHNOLOGY (SHANDONG ACADEMY OF SCIENCES)
Filing Date
2026-03-26
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

In CPU-GPU heterogeneous systems, existing technologies struggle to dynamically adapt to computational characteristics, leading to an imbalance in load between the CPU and GPU, which affects overall computational efficiency and resource utilization.

Method used

An adaptive partitioning decision method is adopted. Through a closed-loop control system based on real-time performance feedback and historical load trends, the task allocation ratio between CPU and GPU is dynamically adjusted. The load is adjusted using a comprehensive scheduling metric. Multi-feature weighted fusion is performed by combining oscillation exploration factor, convergence acceleration factor and load distribution factor to achieve adaptive optimal load allocation.

Benefits of technology

It significantly improves the load balancing of CPU-GPU heterogeneous systems, enhances overall computing performance and resource utilization, optimizes task allocation between CPU and GPU, and improves program execution efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121918971B_ABST
    Figure CN121918971B_ABST
Patent Text Reader

Abstract

The present application relates to a kind of adaptive partition decision task scheduling method and system based on integrated computing architecture, belong to the technical field of electronic information.It includes: step 1: using adaptive partition decision to allocate CPU data volume reasonably;It refers to: through real-time system state monitoring, historical trend analysis and intelligent feedback regulation, the adaptive optimal allocation of work load is realized in CPU-GPU heterogeneous computing architecture;Step 2: execute CPU-GPU task scheduling.The present application method is used to optimize the work load allocation in the environment of heterogeneous processor (for example, central processing unit and graphics processing unit), through adaptive task mapping and scheduling strategy, effectively bridge the architecture and performance difference between different processing units, so as to achieve efficient load balancing at system level, finally significantly improve overall computing performance and resource utilization.
Need to check novelty before this filing date? Find Prior Art