Computing power-power cross-time and space collaborative scheduling method, system and storage medium

By using multidimensional perception and heterogeneous mapping models, combined with user intent and a two-level buffer queue, spatiotemporal collaborative scheduling of computing power and power systems was achieved, solving the problems of power consumption and latency of heterogeneous hardware in the scheduling system, and realizing refined resource scheduling and cost optimization.

CN122363869APending Publication Date: 2026-07-10HEFEI ZHONGKE LEINAO INTELLIGENCE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HEFEI ZHONGKE LEINAO INTELLIGENCE TECH CO LTD
Filing Date
2026-06-10
Publication Date
2026-07-10

AI Technical Summary

Technical Problem

Existing scheduling systems cannot perceive the physical power consumption characteristics of heterogeneous hardware, ignore the differences in latency sensitivity of business scenarios, and the scheduling strategies are limited to a single computing center or static strategies, resulting in a lack of coordination between computing power and power system resource scheduling.

Method used

By sensing the state of computing power and power systems in multiple dimensions, a heterogeneous mapping model is constructed. User intent perception and a two-level buffer queue mechanism are introduced to achieve spatiotemporal collaborative decision-making and dynamically match power and computing resources.

Benefits of technology

It enables unified management of heterogeneous computing power clusters and integration with power trading, fine-tunes computing load, ensures service quality and reduces costs, and solves cross-domain latency bottlenecks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122363869A_ABST
    Figure CN122363869A_ABST
Patent Text Reader

Abstract

This invention discloses a method, system, and storage medium for cross-temporal and spatial collaborative scheduling of computing power and power. The method, based on embodiments of this invention, constructs a heterogeneous mapping model between computing power and power to address the semantic gap between logical computing resources and physical power load, thereby digitizing regulation capabilities. Furthermore, by introducing user intent awareness and a two-level buffer queue mechanism, it resolves the multi-objective game between computing power cost and service quality while ensuring Service Level Agreements (SLAs). Finally, through cross-domain mirroring preheating and cost-aware routing, it addresses the latency bottleneck of computing tasks "following green" in a wide-area space, achieving dynamic spatiotemporal matching of computing power flow and power flow. Thus, it unifies the management of multiple heterogeneous computing power clusters downwards and connects upwards to the power trading and scheduling system, enabling proactive and refined adjustment of computing load.
Need to check novelty before this filing date? Find Prior Art