NoC Data Packet Routing via Intermediate Nodes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In network-on-chip (NoC) systems, data packet transmission efficiency is low due to congestion when multiple packets need to be transmitted between nodes, as they are sent sequentially, leading to bottlenecks and increased latency.

Innovation Solution

A method where intermediate routing nodes are determined based on the source and destination nodes, as well as the data transmission state of each node, and identification information is stored in the data packet's head flit to enable sequential transmission through these nodes using the XY routing algorithm, bypassing congested paths.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If data packets are transmitted sequentially between connected nodes, then transmission order is maintained, but transmission efficiency decreases and congestion occurs

Engineering Contradiction:
Improvedata packet transmission efficiencyVSAvoidtransmission latency
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent segments the data transmission path by dividing it into multiple independent segments between connected nodes. Each segment can transmit data packets independently, allowing parallel transmission across different segments of the same logical path. This segmentation enables multiple packets to be transmitted simultaneously through different intermediate routing nodes, thereby improving overall transmission efficiency and reducing congestion.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces an additional dimension to the transmission path by inserting intermediate routing nodes between source and destination nodes. Instead of direct sequential transmission, data packets can route through multiple intermediate nodes in parallel, creating multiple dimensional paths. This allows simultaneous transmission of multiple packets through different intermediate routing nodes, effectively adding a spatial dimension to the transmission process and reducing latency.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If intermediate routing nodes are inserted to enable parallel transmission, then transmission efficiency improves, but routing complexity increases

Engineering Contradiction:
Improvedata packet transmission efficiencyVSAvoidrouting management complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-determining and storing optimal intermediate routing nodes before actual data transmission occurs. The scheduling management unit calculates and identifies the most efficient intermediate routing nodes in advance, storing this routing information for future use. When data packets need to be transmitted, the pre-computed routing paths are directly applied, eliminating the need for complex real-time routing decisions and reducing routing management complexity while maintaining high transmission efficiency.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The scheduling management unit acts as an intermediary between the source node and intermediate routing nodes. It centralizes the routing decision-making process, managing and coordinating the selection of intermediate routing nodes without requiring complex interactions between multiple nodes. This intermediary approach simplifies routing management by consolidating control functions in a single management unit that handles path optimization and node selection.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS20240244013A1Method and scheduling management unit for transmitting data packet
Publication Date: 2024.07.18 ALIBABA (CHINA) CO LTD
  • US20240244013A1 patent drawing
  • US20240244013A1 patent drawing
  • US20240244013A1 patent drawing

AI summary

Embodiments of this disclosure provide a data packet transmission method, a scheduling management unit, a chip, and a graphic card. The data packet transmission method includes: determining a source node and a destination node of a data packet to be transmitted; determining at least one intermediate routing node corresponding to the data packet to be transmitted based on the source node and the destination node of the data packet to be transmitted and a data transmission state of each node in a network on chip (NoC); and transmitting identification information of the at least one intermediate routing node to the source node of the data packet to be transmitted.