Parallel Processor Data Transmission via Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In high-performance computing applications like molecular dynamics simulations, existing parallel data processing technologies face inefficiencies due to excessive data synchronization and communication overhead between processor modules, which can slow down processing speed and increase resource utilization.

Innovation Solution

The proposed solution involves a parallel processing apparatus and method where processor modules determine whether to transmit operation result data based on differences and thresholds, using prediction algorithms like Taylor series approximation to reduce data transmission, and compressing data to minimize communication overhead, allowing for synchronized operation and efficient data restoration.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If processors perform parallel data processing with frequent synchronization, then processing accuracy and consistency are improved, but communication overhead and processing time increase

Engineering Contradiction:

Improvedata consistencyVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent implements periodic data transmission where the first processor module sends operation result data to the second processor module at specific time intervals rather than continuously. This periodic action reduces communication overhead while maintaining data consistency, as the second processor module can predict intermediate data between transmissions using prediction algorithms.

Inventive Principle:

Principle #19Periodic action

Solution Approach 2:

The second processor module performs preliminary prediction of operation result data using prediction algorithms (such as Taylor series approximation) based on previously received data. This preliminary action allows the module to prepare expected values in advance, reducing the need for frequent synchronization and communication with the first processor module.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If processors exchange intermediate data frequently, then processing accuracy is improved, but communication overhead increases

Engineering Contradiction:

Improveprediction accuracyVSAvoidcommunication overhead

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces a prediction algorithm as an intermediary mechanism between the two processor modules. Instead of directly exchanging all intermediate data, the second processor module uses the prediction algorithm to generate expected values based on previously received operation result data. This intermediary approach reduces communication overhead while maintaining sufficient accuracy for parallel processing.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent changes the parameter of data transmission from continuous to intermittent based on threshold comparisons. The first processor module compares the difference between current and predicted operation result data, and only transmits data when the difference exceeds a predetermined threshold. This parameter change optimizes communication overhead by transmitting only necessary data.

Inventive Principle:

Principle #35Parameter changes

3Loss of information

If all operation result data are transmitted, then data availability is improved, but data transmission volume increases

Engineering Contradiction:

Improvedata availabilityVSAvoiddata transmission volume

Core Design Contradiction:

Loss of informationVSQuantity of substance

Solution Approach 1:

The patent extracts and transmits only the necessary operation result data that exceed a predetermined threshold difference from predicted values, rather than transmitting all operation result data. This extraction approach maintains data availability for accurate parallel processing while significantly reducing data transmission volume between processor modules.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20230229396A1Apparatus and method with parallel data processing

Publication Date: 2023.07.20 SAMSUNG ELECTRONICS CO LTD

AI summary

An apparatus with parallel processing includes: a first processor module; and a second processor module configured to perform parallel processing in synchronization with the first processor module, wherein the first processor module is configured to: determine first operation result data using an operation process in a first time interval; transmit the first operation result data to the second processor module; determine second operation result data using the operation process in a second time interval; and determine whether to transmit the second operation result data to the second processor module, and wherein the second processor module is configured to determine second prediction result data corresponding to the second operation result data based on the first operation result data received from the first processor module and a prediction process in response to the first processor module determining not to transmit the second operation result data to the second processor module.