LSoE Frame Processing Engine for Distributed Virtual Bridge Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Server computers experience transmission bottlenecks and delays when forwarding data frames through centralized switches due to contention for routing resources, leading to inefficiencies in data access and increased hardware costs and latency.
Innovation Solution
A Load Store over Ethernet (LSoE) frame processing engine equipped with a repeat function and a repeat with strobe function, coupled with an auto-increment function of an indirect register facility, enables efficient data transfer from a controlling bridge to a distributed virtual bridge, reducing data loading on computer data paths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If indirect register interface with control register and data registers is used to access large tables, then data access capability is improved, but frame size increases causing competition with normal Ethernet traffic for buffering resources and interconnect link bandwidth
Solution Approach 1:
The patent segments the register access operation into two distinct phases: a control phase that sets up the indirect register interface parameters, and a data phase that transfers actual table data. This segmentation allows the system to use smaller control frames for setup and then efficiently transfer large data payloads without continuous competition with normal Ethernet traffic, resolving the contradiction between data access capability and frame size complexity.
Solution Approach 2:
The patent implements preliminary action by first establishing the indirect register interface configuration before actual data table access begins. The control register is programmed with the necessary addressing and control information in advance, allowing subsequent data transfers to proceed more efficiently without repeated setup overhead, thus reducing the impact of large frames on network resources.
2Ease of operation
If centralized switches are used for routing, then space and connectivity considerations are satisfied, but transmission bottlenecks and delays occur due to contention for routing resources
Solution Approach 1:
The patent extracts the routing resource contention problem by implementing a distributed virtual bridge architecture where bridge elements independently manage their own forwarding decisions. This removes the centralized switch bottleneck by distributing the routing function across multiple independent bridge elements, allowing simultaneous data transfers without centralized resource contention while maintaining ease of operation through standardized Ethernet interfaces.
Solution Approach 2:
The patent transitions from a single-point centralized routing model to a multi-point distributed routing model, adding the dimension of distribution across multiple bridge elements. This dimensional change allows parallel data transfer paths to emerge, eliminating the transmission bottleneck at centralized switches while preserving the operational simplicity of Ethernet connectivity through each bridge element.
3Productivity
If the number of switches and connections is increased to accommodate additional traffic, then traffic capacity is improved, but configuration and management challenges increase and hardware costs and latency rise
Solution Approach 1:
The patent implements universality by designing bridge elements that can perform multiple functions: data forwarding, register access, and configuration management through a unified distributed virtual bridge interface. This multi-functionality eliminates the need for separate dedicated management infrastructure, allowing increased traffic capacity through software-defined virtual bridges rather than proportional increases in physical switches and connections, thus reducing configuration and management complexity.
Data Source
AI summary
A system and method for efficient transfer of data from a controlling bridge to a register of a distributed bridge element. A Load Store over Ethernet (LSoE) frame processing engine (FPE) is equipped with a repeat and a repeat with strobe function that, when coupled with an auto-increment function of an indirect register facility, allows a distributed virtual bridge to move data payload more efficiently which decreases the data loading on the computers data paths used for other data transfers.