Data communications in a parallel active messaging interface of a parallel computer

• US 8,584,141 B2
• Filed: 01/17/2011
• Issued: 11/12/2013
• Est. Priority Date: 01/17/2011
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Archer, Charles J., Blocksome, Michael A., Ratterman, Joseph D., Smith, Brian E.
• Primary Examiner(s): CAO, DIEM K
• Application Number:
  US13/007,860
• Publication Number:
  US 20120185873A1
• Time in Patent Office: 1,030 Days
• Field of Search: 719/313, 709/208
• US Class Current: 719/313
• CPC Class Codes: G06F 15/17318 Parallel communications techniques, e.g. gather, scatter, reduce, roadcast, multicast, all to all
• CPC Class Codes: G06F 9/546 Message passing systems or structures, e.g. queues

Abstract

Data communications in a parallel active messaging interface (‘PAMI’) of a parallel computer composed of compute nodes that execute a parallel application, each compute node including application processors that execute the parallel application and at least one management processor dedicated to gathering information regarding data communications. The PAMI is composed of data communications endpoints, each endpoint composed of a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources. Embodiments function by gathering call site statistics describing data communications resulting from execution of data communications instructions and identifying in dependence upon the call cite statistics a data communications algorithm for use in executing a data communications instruction at a call site in the parallel application.

Claims

1. A method of data communications in a parallel active messaging interface (‘

PAMI’

) of a parallel computer, the parallel computer comprising;

a plurality of compute nodes that execute a parallel application, each compute node comprising a plurality of computer processors including a plurality of application processors that execute the parallel application and at least one management processor dedicated to gathering information regarding data communications, and

the PAMI, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources,

the method comprising;

gathering, by a particular management processor of a particular compute node, call site statistics describing data communications resulting from execution through the PAMI of data communications instructions including gathering the call site statistics into storage administered by the particular management processor; and

identifying through the PAMI by the parallel application in dependence upon the call cite statistics a data communications algorithm for use in executing through the PAMI a data communications instruction at a call site in the parallel application including querying the particular management processor.

2. The method of claim 1 wherein:

gathering call site statistics further comprises gathering the call site statistics into storage available to all of the application processors of the particular compute node; and

identifying a data communications algorithm further comprises querying the PAMI in dependence upon the storage available to all of the application processors of the particular compute node.

3. The method of claim 1 further comprising:

establishing in the PAMI, available to the application processors and to the particular management processor, a list of data communications algorithms; and

sorting the list by the particular management processor in dependence upon the gathered call cite statistics so that the current best algorithm according to the statistics is at the top of the list;

wherein identifying a data communications algorithm further comprises querying the PAMI to identify the algorithm at the top of the list.

4. The method of claim 1 further comprising:

establishing in the PAMI available to the particular management processor a list of data communications algorithms;

sorting the list by the particular management processor in dependence upon the gathered call cite statistics so that the current best algorithm according to the statistics is at the top of the list;

storing by the particular management processor an identifier of the current best algorithm in a single random access memory variable available to the application processors;

wherein identifying a data communications algorithm further comprises querying the PAMI for the stored identifier.

5. The method of claim 1 wherein the call site statistics comprise data communications instruction types, data types, message size, algorithms selected, and performance statistics for each call site.

6. The method of claim 1 wherein:

each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity;

each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and

each task represents a process of execution of the parallel application.

12. The method of claim 1 wherein:

each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity;

each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and

each task represents a process of execution of the parallel application.

7. A parallel computer that implements data communications in a parallel active messaging interface (‘

PAMI’

), the parallel computer comprising;

a plurality of compute nodes that execute a parallel application, each compute node comprising a plurality of computer processors including a plurality of application processors that execute the parallel application and at least one management processor dedicated to gathering information regarding data communications, and

the PAMI, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources,

the computer processors operatively coupled to computer memory having disposed within it computer program instructions that, when executed by the computer processors, cause the parallel computer to function by;

gathering, by a particular management processor of a particular compute node, call site statistics describing data communications resulting from execution through the PAMI of data communications instructions including gathering the call site statistics into storage administered by the particular management processor; and

identifying through the PAMI by the parallel application in dependence upon the call cite statistics a data communications algorithm for use in executing through the PAMI a data communications instruction at a call site in the parallel application including querying the particular management processor.

8. The parallel computer of claim 7 wherein:

gathering call site statistics further comprises gathering the call site statistics into storage available to all of the application processors of the particular compute node; and

identifying a data communications algorithm further comprises querying the PAMI in dependence upon the storage available to all of the application processors of the particular compute node.

9. The parallel computer of claim 7 further comprising computer program instructions that, when executed by the computer processors, cause the parallel computer to function by:

establishing in the PAMI, available to the application processors and to the particular management processor, a list of data communications algorithms; and

sorting the list by the particular management processor in dependence upon the gathered call cite statistics so that the current best algorithm according to the statistics is at the top of the list;

wherein identifying a data communications algorithm further comprises querying the PAMI to identify the algorithm at the top of the list.

10. The parallel computer of claim 7 further comprising computer program instructions that, when executed by the computer processors, cause the parallel computer to function by:

establishing in the PAMI available to the particular management processor a list of data communications algorithms;

sorting the list by the particular management processor in dependence upon the gathered call cite statistics so that the current best algorithm according to the statistics is at the top of the list;

storing by the particular management processor an identifier of the current best algorithm in a single random access memory variable available to the application processors;

wherein identifying a data communications algorithm further comprises querying the PAMI for the stored identifier.

11. The parallel computer of claim 7 wherein the call site statistics comprise data communications instruction types, data types, message size, algorithms selected, and performance statistics for each call site.

13. A computer program product for data communications in a parallel active messaging interface (‘

PAMI’

) of a parallel computer, the parallel computer comprising;

a plurality of compute nodes that execute a parallel application, each compute node comprising a plurality of computer processors including a plurality of application processors that execute the parallel application and at least one management processor dedicated to gathering information regarding data communications, and

the PAMI, the PAMI comprising data communications endpoints, each endpoint comprising a specification of data communications parameters for a thread of execution on a compute node, including specifications of a client, a context, and a task, the compute nodes and the endpoints coupled for data communications through the PAMI and through data communications resources,

the computer program product disposed upon a computer readable storage medium, wherein the computer readable storage medium is not a signal, the computer program product comprising computer program instructions that, when installed and executed, cause the parallel computer to function by;

gathering, by a particular management processor of a particular compute node, call site statistics describing data communications resulting from execution through the PAMI of data communications instructions including gathering the call site statistics into storage administered by the particular management processor; and

identifying through the PAMI by the parallel application in dependence upon the call cite statistics a data communications algorithm for use in executing through the PAMI a data communications instruction at a call site in the parallel application including querying the particular management processor.

14. The computer program product of claim 13 wherein:

gathering call site statistics further comprises gathering the call site statistics into storage available to all of the application processors of the particular compute node; and

identifying a data communications algorithm further comprises querying the PAMI in dependence upon the storage available to all of the application processors of the particular compute node.

15. The computer program product of claim 13 further comprising computer program instructions that, when installed and executed, cause the parallel computer to function by:

establishing in the PAMI, available to the application processors and to the particular management processor, a list of data communications algorithms; and

sorting the list by the particular management processor in dependence upon the gathered call cite statistics so that the current best algorithm according to the statistics is at the top of the list;

wherein identifying a data communications algorithm further comprises querying the PAMI to identify the algorithm at the top of the list.

16. The computer program product of claim 13 further comprising computer program instructions that, when installed and executed, cause the parallel computer to function by:

establishing in the PAMI available to the particular management processor a list of data communications algorithms;

sorting the list by the particular management processor in dependence upon the gathered call cite statistics so that the current best algorithm according to the statistics is at the top of the list;

storing by the management processor an identifier of the current best algorithm in a single random access memory variable available to the application processors;

wherein identifying a data communications algorithm further comprises querying the PAMI for the stored identifier.

17. The computer program product of claim 13 wherein:

each client comprises a collection of data communications resources dedicated to the exclusive use of an application-level data processing entity;

each context comprises a subset of the collection of data processing resources of a client, context functions, and a work queue of data transfer instructions to be performed by use of the subset through the context functions operated by an assigned thread of execution; and

each task represents a process of execution of the parallel application.