Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Expedited Mathematical Processing on LAN Machines

IP.com Disclosure Number: IPCOM000110826D
Original Publication Date: 1994-Jan-01
Included in the Prior Art Database: 2005-Mar-26
Document File: 2 page(s) / 54K

Publishing Venue

IBM

Related People

Johnson, WJ: AUTHOR [+2]

Abstract

In scientific applications, the user frequently writes programs to solve very complex algebraic or statistical equations. These computations are CPU-intensive, and sometimes they take a considerable amount of CPU time. This is often due to the inclusion of functions such as logarithms, exponentiation, or factorial calculations. A method is needed to speed up the performance of these types of calculations in a multi-processing environment, capable of using multiple hardware processors in a LAN environment.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 100% of the total text.

Expedited Mathematical Processing on LAN Machines

      In scientific applications, the user frequently writes programs
to solve very complex algebraic or statistical equations.  These
computations are CPU-intensive, and sometimes they take a
considerable amount of CPU time.  This is often due to the inclusion
of functions such as logarithms, exponentiation, or factorial
calculations.  A method is needed to speed up the performance of
these types of calculations in a multi-processing environment,
capable of using multiple hardware processors in a LAN environment.

      This system provides a method for a user on a LAN requester to
enable the dividing of complex mathematical equations into smaller
logical units, then spinning new threads or processes on other LAN
requesters to solve several parts of the equation simultaneously.