Location-Based NoC Subtopologies for SoC Wire Delay Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

As microprocessors handle larger and more complex workloads, traditional network-on-chip (NoC) topologies face limitations due to long wire delays and dense logic, which can lead to performance bottlenecks and timing issues in system-on-chip (SoC) designs.

Innovation Solution

The implementation of a location-based NoC interface with subtopologies, where the NoC topology is divided into subtopologies based on the physical location of logic blocks, each including at least one router. This approach optimizes the placement and coupling of subtopologies using communications protocols, enabling efficient data transfer between logic blocks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a single NoC topology is used to connect all logic blocks, then the communication infrastructure is simple, but long wire delays and performance bottlenecks occur

Engineering Contradiction:
ImproveNoC topology structureVSAvoidwire delay
Core Design Contradiction:
Device complexityVSLoss of time

Solution Approach 1:

The patent divides the single NoC topology into multiple subtopologies based on the physical location of logic blocks. Each subtopology serves a specific region or functional unit, reducing the maximum communication distance within each subtopology and thereby decreasing wire delays while maintaining overall system connectivity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a hierarchical dimension to the NoC architecture by organizing subtopologies into groups and levels. This multi-level hierarchy allows communication to occur at different scales - local communication within subtopologies and global communication between them - effectively reducing the time penalty of long-distance wires through dimensional organization.

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

2Area of stationary object

If logic blocks are densely packed to increase chip capacity, then area efficiency improves, but communication paths become congested and timing issues arise

Engineering Contradiction:
Improvechip area utilizationVSAvoidtiming requirement
Core Design Contradiction:
Area of stationary objectVSReliability

Solution Approach 1:

The patent applies local quality by creating subtopologies with specialized characteristics suited to their specific regions or functional units. Each subtopology can be optimized for its local communication patterns and timing requirements, allowing dense packing while maintaining reliable timing through location-aware design.

Inventive Principle:
Principle #3Local quality

3Speed

If the NoC topology is optimized for current performance, then communication speed improves, but adaptability to future scaling and new configurations is reduced

Engineering Contradiction:
Improvecommunication speedVSAvoidscalability
Core Design Contradiction:
SpeedVSAdaptability or versatility

Solution Approach 1:

The patent creates a dynamic and flexible NoC architecture where subtopologies can be independently configured, added, or modified based on evolving requirements. The hierarchical structure and location-based organization allow the system to adapt to future scaling and new configurations while maintaining optimized communication paths for current performance needs.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS20250028889A1Location-based noc interface with subtopologies
Publication Date: 2025.01.23 SIGNATURE IP CORP
  • US20250028889A1 patent drawing
  • US20250028889A1 patent drawing
  • US20250028889A1 patent drawing

AI summary

Disclosed embodiments provide techniques for a networking with a network-on-chip (NoC) interface with subtopologies. A system-on-chip (SoC) is accessed. The SoC includes a plurality of logic blocks. A NoC topology is created. The NoC topology includes one or more subtopologies. The one or more subtopologies are based on a physical location of the plurality of logic blocks. Each subtopology includes at least one router. A location of the one or more subtopologies is optimized. The one or more subtopologies are coupled based on one or more communications protocols. A protocol running on the plurality of logic blocks is translated to the one or more communications protocols. Data is sent from a sending subtopology within the one or more subtopologies to a receiving subtopology within the one or more subtopologies.