Stream Context Switching for Low-Latency Data Transformation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data storage servers face challenges in bulk data transfer, particularly in streaming video applications, due to high latency caused by inefficient context switching and data transformation processes, which are affected by variations in input arrival rates and require significant CPU cycles, leading to reduced quality of transformations such as compression and encryption.

Innovation Solution

Implementing a context switching mechanism where a processing core generates and stores context data including transformation history for each data stream, allowing for coordinated data transfer between processing cores and context memory, enabling improved compression and encryption by leveraging previous context information for subsequent data packets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If context switching is performed frequently to handle multiple data streams, then adaptability to varying input arrival rates is improved, but latency increases due to repeated initialization of transformation engines

Engineering Contradiction:

Improveadaptability to input arrival ratesVSAvoidlatency

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-loading and caching context data in dedicated context memory before it is needed. When a data stream arrives, the corresponding context data is already available in memory, eliminating the need for real-time initialization and reducing latency while maintaining adaptability to varying input rates

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically switches between different context data sets in context memory based on the arriving data stream characteristics. The context control module dynamically selects and loads the appropriate context data without fixed timing, allowing the system to adapt to varying input arrival rates while avoiding repeated full initializations

Inventive Principle:

Principle #15Dynamics

2Productivity

If data transformation is performed on each packet independently, then processing speed is improved, but quality of transformation deteriorates due to loss of state information

Engineering Contradiction:

Improveprocessing speedVSAvoidquality of data transformation

Core Design Contradiction:

ProductivityVSManufacturing precision

Solution Approach 1:

The system segments the data stream into packets for parallel processing while maintaining separate context data for each stream in context memory. Each packet can be processed independently at high speed, but the context data preserves the state information needed to maintain transformation quality across packets of the same stream

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Context data acts as an intermediary between packets of the same data stream. It carries the state information that would otherwise be lost between independent packet transformations, enabling both high processing speed through parallel packet handling and high transformation quality through preserved state context

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If multiple protocol stacks and processing modules are used for data transformation, then versatility of data handling is improved, but device complexity increases leading to more CPU cycles and higher latency

Engineering Contradiction:

Improveversatility of data handlingVSAvoidcomplexity of processing architecture

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The context control module serves multiple functions: it manages context data storage, retrieves context data, switches between different data streams, and coordinates with the processing core. This single multi-functional module replaces what would otherwise require multiple separate processing modules, reducing device complexity while maintaining versatility

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system merges the context management functionality into a unified context control module that handles all state information for multiple data streams. Context memory consolidates the state data that would otherwise be scattered across multiple processing modules, reducing overall system complexity while preserving the ability to handle diverse data transformation requirements

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS8868674B2Streaming and bulk data transfer transformation with context switching

Publication Date: 2014.10.21 EXAR CORP

AI summary

In described embodiments, processing of a data stream, such as a packet stream or flow, associated with data streaming is improved by context switching that employs context history. For each data stream that is transformed through processing, a context is maintained that comprises state information and includes a history and state information that enables the transformation for the data stream. Processing for the data transformation examines currently arriving data and then processes the data based on the context data and previously known context information for the data stream from the history stored in memory.