一种片上网络系统拓扑生成方法

By generating irregular single-path and multi-path topologies and optimizing transmission delay and area using multi-objective functions, the problem of complex design and low resource utilization in existing on-chip network topology generation methods is solved, achieving high-efficiency communication and low-power multi-core processor performance improvement.

CN117931727BActive Publication Date: 2026-07-17FUDAN UNIVERSITY

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
FUDAN UNIVERSITY
Filing Date
2024-01-17
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing on-chip network topology generation methods rely on expert experience, resulting in complex and time-consuming designs. They cannot adapt to the increasing scale of multi-core processors and the diversification of application requirements, leading to high communication latency, low bandwidth utilization, and traditional regular topology structures that cannot fully utilize the parallel computing capabilities of multi-core processors.

Method used

The range of router numbers is calculated using a saturation growth method, generating irregular single-path and multi-path topologies. Transmission delay and area are optimized through multi-objective functions to generate topologies that adapt to different application requirements.

Benefits of technology

It improves the performance of multi-core processors, reduces power consumption, enhances system flexibility and communication efficiency, and solves the problems of performance bottlenecks and low resource utilization in traditional methods.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117931727B_ABST
    Figure CN117931727B_ABST
Patent Text Reader

Abstract

本发明属于集成电路设计技术领域,具体为一种片上网络系统拓扑生成方法。本发明的片上网络拓扑生成方法,包括:利用任务节点数和路由器端口数,采用饱和增长方式,计算出最终拓扑结构中的路由器数量范围(rmin,rmax),并生成非规则的单通路拓扑结构;为单通路连接的路由器增加与相邻路由器相连的链路,生成具有多种传输路径的多通路拓扑结构;在满足系统约束的条件下,将片上网络中延迟和面积作为性能指标,定义拓扑生成过程中的多目标函数,选取目标函数值较低、性能更优的拓扑结构;本发明可充分利用多核处理器中的资源和并行计算能力,同时解决了单目标约束条件生成的拓扑结构整体性能不均衡的问题。
Need to check novelty before this filing date? Find Prior Art