Single NIC Hardware Offload Engine for Multi-Path Communication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing data communication systems require multiple Network Interface Cards (NICs) with hardware offload engines to ensure reliability, leading to increased costs due to the need for as many NICs as communication paths, which is inefficient and costly.
Innovation Solution
A data communication apparatus that utilizes a single hardware protocol processing unit, such as a TCP/IP Offload Engine (TOE), to manage multiple communication paths by storing identifiers for each network interface, allowing the system to perform protocol processing and route data efficiently across multiple networks without the need for multiple NICs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple NICs with hardware offload engines are used to ensure reliability, then reliability is improved, but device complexity and cost increase
Solution Approach 1:
A single NIC with hardware offload engine is designed to perform multiple functions by supporting multiple communication paths through software-based protocol processing. The hardware offload engine processes TCP/IP protocols for different networks, enabling one NIC to replace what would traditionally require multiple NICs to achieve similar reliability through path diversity.
Solution Approach 2:
The patent introduces a software-based protocol processing layer that acts as an intermediary between the hardware offload engine and multiple network paths. This software layer manages protocol processing for different networks, allowing the hardware to focus on high-speed data transfer while the software coordinates reliability across multiple paths.
2Reliability
If multiple NICs with hardware offload engines are used to ensure reliability, then reliability is improved, but cost increases
Solution Approach 1:
The hardware offload engine is designed with universal functionality to handle protocol processing for multiple networks through a single NIC. This multi-functionality eliminates the need to purchase and deploy multiple expensive NICs, reducing overall system cost while maintaining the reliability benefits of multiple communication paths.
Solution Approach 2:
The patent merges the functionality of multiple NICs into a single NIC by combining multiple protocol processing functions within one hardware offload engine. This consolidation reduces the total number of hardware components needed, thereby reducing cost while preserving the reliability advantages through software-coordinated multi-path communication.
3Adaptability or versatility
If software processes TCP/IP protocols, then flexibility is maintained, but processing load increases and speed decreases
Solution Approach 1:
The patent segments protocol processing into two distinct layers: hardware-based TCP/IP protocol processing handled by the hardware offload engine for high-speed data transfer, and software-based higher-layer protocol processing for flexibility and adaptability. This segmentation allows each layer to optimize for its specific function without compromising the other.
Solution Approach 2:
The patent replaces pure software-based TCP/IP protocol processing with a hardware-based hardware offload engine that performs protocol processing in parallel to software. This substitution maintains protocol processing flexibility while dramatically improving communication speed by moving critical protocol handling from software to hardware.
Data Source
AI summary
In one embodiment, a data communication apparatus includes a plurality of communications units for performing transmission and reception with the plurality of different networks, a first storage unit which stores a first identifier and a second identifier, a first processing unit which performs network protocol processing with hardware, and a second processing unit which performs network protocol processing with software. The first identifier indicates a communication unit which receives a frame, and the second identifier indicates a communication unit which transmits a frame. The first processing unit stores the first identifier in association with a reception frame in the first storage unit, and the second processing unit reads the first identifier and the reception frame from the first storage unit, generates a transmission frame as a response to the reception frame, stores the first identifier as the second identifier in association with the transmission frame in the first storage unit.