Rackmount PCIe Switch Assembly with Redundant Cross-Link Ports

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Traditional PCIe implementations in networked storage and computing systems lack redundancy, leading to potential failures and disruptions in data processing and storage, especially in high-density rack-mounted environments where point-to-point host-device architectures are vulnerable to single-point failures.

Innovation Solution

The development of a rackmount PCIe switch assembly with redundancy cross-link ports and a control processor that forms a cluster interconnect PCIe fabric, enabling failover traffic management between multiple switch assemblies, ensuring continuous operation even if one switch fails.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If point-to-point host-device PCIe architecture is used, then device connectivity is achieved, but system reliability deteriorates due to single-point failures

Engineering Contradiction:
Improvesystem reliabilityVSAvoidarchitecture complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system segments PCIe connectivity into multiple independent paths through switch assemblies. Each switch assembly divides the monolithic point-to-point architecture into modular units with multiple ports, allowing traffic to be routed through different paths. This segmentation enables failover capabilities where if one path fails, traffic can be redirected through alternative paths, thereby improving reliability without requiring a complete system redesign.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The PCIe switch assemblies implement dynamic path selection and failover mechanisms. The system can dynamically reroute traffic based on the operational status of different paths, transitioning from static point-to-point connections to dynamic multi-path routing. This dynamic capability allows the system to adapt to failures in real-time, maintaining connectivity and improving reliability while managing complexity through intelligent control.

Inventive Principle:
Principle #15Dynamics

2Reliability

If redundancy cross-link ports are added to switch assemblies, then failover capability is improved, but device complexity increases

Engineering Contradiction:
Improvefailover capabilityVSAvoidswitch assembly complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

Multiple switch assemblies are merged into a unified PCIe fabric through cross-link ports. Rather than treating each switch as an isolated complex unit, the patent combines them into a coordinated system where redundancy is achieved through inter-assembly linkages. This merging approach distributes the complexity across multiple standardized units rather than concentrating it in single complex switches, improving failover capability while managing overall system complexity through modularity.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The cross-link ports serve multiple functions: they provide redundancy for failover, enable inter-assembly communication, and facilitate load balancing. This multi-functionality reduces the need for dedicated components for each purpose, thereby improving reliability through redundancy while minimizing the increase in device complexity by making existing components serve multiple roles.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Reliability

If cluster interconnect PCIe fabric is formed, then traffic failover is enabled, but system complexity increases

Engineering Contradiction:
Improvetraffic failoverVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The PCIe switch assemblies are nested within a larger cluster interconnect fabric, with each assembly containing its own internal switching fabric and the overall system forming a hierarchical structure. This nesting allows traffic failover to be managed at multiple levels - internally within each switch assembly and externally across the cluster fabric. The hierarchical nesting simplifies complexity management by organizing functions at appropriate levels rather than requiring a single monolithic complex system.

Inventive Principle:
Principle #7Nested doll (Nesting)

Data Source

PatentUS10747701B2Rackmount switch devices for peripheral component interconnect express (PCIe) systems
Publication Date: 2020.08.18 LIQID INC
  • US10747701B2 patent drawing
  • US10747701B2 patent drawing
  • US10747701B2 patent drawing

AI summary

Rackmount Peripheral Component Interconnect Express (PCIe) switch assemblies are provided herein. One example PCIe switch assembly includes an enclosure that encases elements of the PCIe switch assembly, a first plurality of PCIe interconnect ports positioned on a front side of the PCIe switch assembly, and a second plurality of PCIe interconnect ports positioned on a rear side of the PCIe switch assembly. One or more redundancy cross-link ports are provided to handle failover traffic with at least another PCIe switch assembly. PCIe switch circuitry is communicatively coupled to the first plurality of PCIe interconnect ports and the second plurality of PCIe interconnect ports that form a cluster interconnect PCIe fabric. A control processor is configured to control operation of at least the PCIe switch circuitry and the one or more redundancy cross-link ports.