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

Optimized, Multi-Tasking, Asynchronous Data Transport Mechanism Across a Synchronous Connection

IP.com Disclosure Number: IPCOM000112883D
Original Publication Date: 1994-Jun-01
Included in the Prior Art Database: 2005-Mar-27
Document File: 2 page(s) / 82K

Publishing Venue

IBM

Related People

Clausen, TL: AUTHOR [+4]

Abstract

A method is disclosed to 'simulate' (and optimize) an asynchronous type data transfer interface over a restricted synchronous hardware connection. It allows for:

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 53% of the total text.

Optimized, Multi-Tasking, Asynchronous Data Transport Mechanism Across
a Synchronous Connection

      A method is disclosed to 'simulate' (and optimize) an
asynchronous type data transfer interface over a restricted
synchronous hardware connection.  It allows for:

o   Distribution of function between a primary processor and a
    secondary processor

o   Moving data from many tasks across a narrow synchronous
    interface, equitably and efficiently

o   Management of asynchronous data requests to prevent over-runs of
    the secondary (or primary) processor buffer allocations

o   Flexibility for data size, parallel task data requests, and
    strict requirements on data transfer time (from task initiation
    to delivery).

The Figure depicts the overall concepts used in this invention.

      This disclosure solves a data passing bottleneck for
synchronous data transfer mechanisms (between processors) with
restrictions on the number of outstanding operations in either
direction at any one time.  It is possible to tie up these mechanisms
with large amounts of data directed only to a few specific tasks
(devices) while starving out others (or locking out device transfers
back).

      Distribution of function from a single processor to separate
processors requires an environment to allow tight coupling between
host and secondary processor tasks, an open, free flowing data
connection, and timely transfers on data requests.

The "key elements" utilized in the disclosed method are:

1.  Multiplexing data from all tasks into 'single' data transfer
    requests initiated via:

    a.  'periodic' timer intervals  <or>
    b.  priority 'immediate' basis

2.  Initiator 'Pacing' of data to/from individual processors to
    control processor 'data buffer overruns'

    o   'self limiting' data sending mechanism

      Item 1 is self-explanatory in that outstanding data requests
from all executing tasks are bundled...