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

Different Forms of Specificity within Homogeneous Workloads

IP.com Disclosure Number: IPCOM000106555D
Original Publication Date: 1993-Nov-01
Included in the Prior Art Database: 2005-Mar-21
Document File: 2 page(s) / 94K

Publishing Venue

IBM

Related People

Rechtschaffen, R: AUTHOR [+2]

Abstract

The generality of the definition of specificity of an execution sequence with a single entry point allows for a broader definition involving multiple entry points to be exploited. The manner in which this is done is disclosed.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Different Forms of Specificity within Homogeneous Workloads

      The generality of the definition of specificity of an execution
sequence with a single entry point allows for a broader definition
involving multiple entry points to be exploited.  The manner in which
this is done is disclosed.

      An illustration of the environments in which specialized
processors can be dynamically created and used to advantage in a
homogeneous workload is given in the following schematic:

     *----------*----------*----------*----------*

     |    S1    |    S2    |    S3    |   S4     |

     *----------*----------*----------*----------*
where S1, S2, ....  are stages of processing of the transaction using
different programs.

      If such transaction processing is done within a uniprocessor,
or done in its entirety within processors of a multiprocessor, the
advantages of isolating the processing of individual stages to
separate processors can not be realized.  If we consider a pipeline
of such transaction processing within a set of dedicated processors,
as shown below, the advantages of isolating a stage to a processor
can be realized.

If each stage of processing, S sub i can be associated with a unique
entry point, EP sub i, that triggers a processor switch to a
specialized processor, P sub i, then a high-level processor
pipelining of the workload can be realized.

      The overall characteristics of such a specialization, as
triggered by entry points that had suitable characteristics, is:

o   The number of processors and the length of each stage can be
    determined by the I-CACHE of the processors.  The length of the
    stage assigned to a specialized processor being the amount of
    execution that can be contained in a prescribed number of cache
    lines.

o   The I-CACHE load to handle S sub i on P sub i will contain all
    the cache lines needed to handle the variation associated with
    branching.

o   It is a simple matter to support a passing of the D-CACHE lines
    as prefetch targets from the TFP (Transferred From Processor) to
    the TOP (Transferred To Processor) based on a queued interface
    and using the value of the registers that transfer the Machine
    State.

o   One should note that the process of specialization can be made
    incremental by first excising S1 ->= P1 and then S2 ->= P2, ...
    etc.

      Different Forms of Specificity within Homogeneous Workloads -
There are several extensions to the use of specificity within
specialized processors.  The particular...