Browse Prior Art Database

Use of Pararell Processing in Data Conversion

IP.com Disclosure Number: IPCOM000013733D
Original Publication Date: 2002-Oct-03
Included in the Prior Art Database: 2003-Jun-18
Document File: 2 page(s) / 65K

Publishing Venue

IBM

Abstract

Disclosed is an architecture for data conversion system which shortens the conversion elapsed time. Here the architecture is intended for use under MP(Multiple Processor) systems of UNIX computers. It uses parallel processing mechanism utilizing UNIX shared memory functionality. The target data in mind is CAD data whose data format vary from one CAD system to another. Data conversion of CAD data from one CAD system to another is generally a time consuming process, taking several tens of minutes in many cases. On the other hand, data conversion system is suitable for multi processing, because:

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

Page 1 of 2

Use of Pararell Processing in Data Conversion

Disclosed is an architecture for data conversion system which shortens the conversion elapsed time.

Here the architecture is intended for use under MP(Multiple Processor) systems of UNIX computers. It uses parallel processing mechanism utilizing UNIX shared memory functionality.

The target data in mind is CAD data whose data format vary from one CAD system to another.

Data conversion of CAD data from one CAD system to another is generally a time consuming process, taking several tens of minutes in many cases. On the other hand, data conversion system is suitable for multi processing, because:
(1)Processing of Input data and output data is generally independent.
(2)Processing of each data entity is generally independent.(For example no relation between one line and another line or curve.) By organizing the conversion system by multiple UNIX processes, the elapsed time can be significantly reduced when multiple

CPUs are available.

The next chart-1 shows the architecture of the conversion system and flow of control. The distinctive characteristics of this architecture are:
(1)Separate input and output side by intermediate data as an interface residing in shared memory.

(Note: The intermediate data is a data format designed freely independent of any CAD systems.)
(2)The input processing and output processing are done in parallel by several processes.

4)Read from shared
memory

Chart -1: Flow of Conversion from CAD Data-A to CAD Data-B

1)Model retrieve: Retrieve each data entity from CAD Data-A.

Done using a subroutine available from CAD-A vender.

2)fork: Unix system function to create a child process.

3)Write to shared memory: Store the intermediate data in

shared memory.

4)Read from shared memory: Read each data entity from

shared memory one by one.

5)Model write: Write each data entity into CAD Data-B.

Done using a subroutine available from CAD-B vender.

...