Composite Node Interconnection for Reduced Chip Count
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current business servers with multi-node interconnection structures require numerous interconnection chips, leading to complex connections, poor reliability, and long delays in inter-node access, which hinder system efficiency.
Innovation Solution
A computer subsystem and system design featuring L composite nodes, each with M basic nodes containing N CPUs and one node controller (NC) where any two CPUs are interconnected, and NCs have routing functions, allowing communication between any two NCs in three hops or less, thereby reducing the need for Xbar interconnection chips and simplifying the connection structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a multi-node interconnection structure with Xbar switches is used, then system scalability is improved, but device complexity increases and reliability deteriorates
Solution Approach 1:
The system is divided into multiple composite nodes, each containing multiple basic nodes with their own NCs. This segmentation allows the system to scale by adding composite nodes while keeping each node's internal structure manageable and independent, reducing overall system complexity.
Solution Approach 2:
The node controller (NC) is designed with multi-functionality, serving as both a local controller within a basic node and as a routing node in the interconnection network. This eliminates the need for separate switching components, simplifying the interconnection structure while maintaining scalability.
2Adaptability or versatility
If a multi-node interconnection structure with Xbar switches is used, then system scalability is improved, but reliability deteriorates
Solution Approach 1:
By segmenting the system into independent composite nodes with self-contained basic nodes and NCs, failure isolation is achieved. A fault in one composite node does not propagate to others, improving system reliability while maintaining scalability through modular addition of nodes.
Solution Approach 2:
The NC acts as an intermediary with routing functionality, providing controlled communication paths between basic nodes and composite nodes. This intermediary structure enables reliable message routing and error handling, ensuring system reliability as the network scales.
3Adaptability or versatility
If a multi-node interconnection structure with Xbar switches is used, then system scalability is improved, but inter-node access delay increases
Solution Approach 1:
The NC performs multiple functions including local control, routing, and direct interconnection management. This multi-functionality eliminates intermediate switching stages, reducing the number of hops and access delay while allowing the system to scale by adding more NCs that can directly interconnect.
Solution Approach 2:
The interconnection structure transitions from a hierarchical multi-stage switch fabric to a more direct topological arrangement where NCs can communicate in three hops or fewer. This dimensional optimization reduces path length and access delay while maintaining system scalability.
Data Source
AI summary
A computer subsystem and a computer system, where the computer subsystem includes L composite nodes (CNs), each CN includes M basic nodes, each basic node includes N central processing units (CPUs) and one node controller (NC). Any two CPUs in each basic node are interconnected. Each CPU in each basic node is connected to the NC in the basic node. The NC in each basic node has a routing function. Any two NCs in the M basic nodes are interconnected. A connection between the L CNs formed through connections between NCs enables communication between any two NCs to be no more than three hops. Hence, the computer subsystem and the computer system can reduce the kinds and the number of interconnection chips, and simplify an interconnection structure of a system, thereby improving reliability of the system.