HPCC and Java -- A Preliminary Report by The Parallel Compiler Runtime Consortium (PCRC)

• Northeast Parallel Architectures Center at Syracuse University: Geoffrey Fox,Bryan Carpenter, Wojtek Furmanski, Don Leskiw, Xiaoming Li
• Cooperating Systems: Marina Chen, James Cowie
• Harvard University: Thomas Cheatham, Amr Fahmy
• Indiana University: Dennis Gannon, Rajeev Raje
• Rice University: Ken Kennedy, Chuck Koelbel, John Mellor-Crummey
• University of Maryland: Joel Saltz, Alan Sussman
• University of Rochester: Wei Li, Michael Scott
• University of Texas: Austin: Jim Browne
  

General Approach:

• ASAP -- people to put on reflector sent to me: reflector is pcrc-java@npac.syr.edu
• February 22 All Initial Material to reflector to be set up by Syracuse
• February 25 Draft 2 deep outline with full writing Assignments
• February 29 2 deep outline with assignments agreed
• Mid April -- all assignments complete -- editing etc. starts
• Draft must be ready for public but friendly reading by CRPC meeting in Mid May
• Based on the manifesto we will call a meeting to decide what areas PCRC will focus on
  

Four areas are available in this document as listed below -- Click to find details!

Initial Table of Suggested Relevant Areas
General Comments are available
Current List of Section Numbers
Full Collection of Initial Text

Initial Table of Suggested Relevant Areas

These are plans made at the Arpa PI meeting february 15 1996

1. What are Suitable Models for Parallel Computing and Java

speculations from Browne

comments fom Indiana

2. Motivating Applications for HPCCJava

distributed simulation etc.

Syracuse and Maryland initial words

3. Lessons from and Relations to HPF/HPC++

Indiana

See also Maryland section 7.4.3

4. Language Enhancements(Annotations) and Classes to support HPCC and Java

Marina Chen, Jim Cowie initial words

Indiana Comment.

5. Architecture (Shared, Distributed Shared, and Distributed) Issues for HPCCJava

Rochester will supply initial words

Indiana Comment

6. Compilers and Java -- Both conventional and HPCC and on client(interpreter) and Server(javac)

Rice will give initial words

Indiana comment

7. High Level Runtime Issues

7.1 Dynamic Communication and Distributed Nameservers

Browne speculations

7.2 Parallel and Conventional CORBA Interface to Java

Gannon Initial words

7.3 Task Parallelism and Java for both Integration(Coarse grain) and fine grain examples

Fortran-M CC++, "channel" classes in Java (see Indiana remark)

Relation to MetaChaos and PCRC Interoperability

Maryland will give initial words -- see sections 7.4.2umd and 7.4.3umd

7.4 HPCC and Compute-Webs (Use of World Wide Web MetaComputer)

HPCC applications as "applets" -- Indiana comment

Maryland Initial words

Boston/CSC/Syracuse may add

7.5 HPCC Java and implications for security

Brokers as interfaces to Java native HPCC related classes

Gannon initial words: see the CORBA discussion in Section 7.2

8 Low Level Runtime Issues

8.1 The Current PCRC Runtime (Regular and Irregular) Implementation in Java

Full (portable) Java version

What is smallest set of PCRC routines that are needed

to be done in native mode (i.e. current C)

Syracuse will provide initial words

8.2 Distributed Memory Versions of Java Runtime and use of HPCC thread packages on SMP or distributed memory to support Java

Rochester and Gannon will get initial words here

back General Comments

• Gannon(27feb96): I just ran across some interesting stuff on the EspressoGrider project http://wwwipd.ira.uka.de/%7Eespresso/Project/ that looks at some interesting extensions to java with a concern about performance, compilers and parallelism. suggst. you all check it out. (one of the people involved Martin Ordersky will be here friday to talk about this work. I will let you know what i learn.
• Note: http://www.communities.com/ - Electric Communities HomePage
• and http://www.communities.com/index/e.html - The E Programming Language
• Summary Presented by Fox at Arpa Meeting:
  • Features of the Java language and runtime
    • Secure, architecture neutral, Object Oriented but simpler than C++
    • runtime modules dynamically downloadable with applet mechanism and portable opcodes (from PC, Mac, to Workstation)
  • Can build on existing work on HPF HPC++ -- especially latter
  • Two basic types of opportunities:
    • Build a modest performance totally portable runtime in Java
    • Use current PCRC runtime as native classes for a (data-parallel) Java
    • What is minimum set of PCRC routines which MUST be native(C)?
  • Applet mechanism naturally incorporates task parallelism -- need to add "channel" class (as in Fortran-M, CC++) to augment "thread" and "socket" classes
    • Need to implement Java runtime in Nexus or similar distributed thread package -- again native Java would be portable but lowish performance
  • Java can be used both on Server and Client side as expect soon good Java based clients and servers
  • See the E language http://www.communities.com/e/index.html
  • or my class notes in general or specifically Java foils http://www.npac.syr.edu/users/gcf/cps616java96 (where foils 115,116 discuss HPJava for class) or
  • http://www.npac.syr.edu/users/gcf/cps616tech96

back Current List of Section Numbers

1. What are Suitable Models for Parallel Computing and Java

• 1.1tex) The Nature of Java Applications
• 1.2tex) Consistency with Current Language Philosophy and Feature Set.
• 1.3tex) Inferences
• 1.4tex) Current Features for Specification of Parallel Execution Behavior
• 1.5tex) A Proposal for Additional Specification of Parallelism for Java
• 1.6tex) What Should NOT be Done
• 1.4ind)Current Features of Java
• 1.5ind)Extended Synchronization in Java
• 1.6ind) Is Data Parallelism relevant for Java:
• 1.7ind)Current Strengths and Weaknesses of Java:

2. Motivating Applications for HPCCJava

• 2.1Syr) Introduction:
• 2.2Syr) Meta-Challenges:
• 2.3Syr) An Example Meta-Challenge: DIS:
• 2.4Syr) Java for DIS:
• 2.5umd) Remote sensing data fusion:
• 2.6umd) Interventional Epidemiology/ Medical crisis management:

3. Lessons from and Relations to HPF/HPC++

See also Maryland section 7.4.3

• 3.1ind) Comparison with HPF:
• 3.2ind) Comparison with HPC++:
• 3.3ind) Operator Overloading:

4. Language Enhancements(Annotations) and Classes to support HPCC and Java

• 4.1csc) Extensions via class libraries.
• 4.2csc) Extensions to the Java language.
• 4.3csc) Implications of Immature Language:
• 4.4csc) Comparison with C++:
• 4.2ind)Implications of Java Philosophy:
• 4.5ind)High Performance Classes:

5. Architecture (Shared, Distributed Shared, and Distributed) Issues for HPCCJava

• 5.1roc) Generate native code for shared-memory architectures:
• 5.2roc) Java vs Bytecode:
• 5.3roc) Locality optimizations for (distributed) share-memory architectures:
• 5.4roc) How to integrate shared-memory environments across a heterogeneous collection of machines and interconnects, respecting the dramatically different latencies and bandwidths:
• 5.5roc) Compiling Java for heterogeneous networks of machines:
• 5.6ind)General Issues:

6. Compilers and Java -- Both conventional and HPCC and on client(interpreter) and Server(javac)

• 6.1rice) Compiled vs. Interpreted
• 6.2rice) Compilation issues
• 6.3rice) Interaction with Runtime
• 6.4rice) Tools Interface
• 6.1ind) Performance Issues:

7. High Level Runtime Issues

7.1 Dynamic Communication and Distributed Nameservers

• 7.1.1tex) Communication Requirements of Java Applications
• 7.1.2tex) The Current Model of Communication
• 7.1.3tex) Desirable Properties of Model of Communication for Java
• 7.1.4tex) Proposed Model of Communication
• 7.1.5tex) Integration/ Implementation Concepts
• 7.1.6tex) Comments

7.2 Parallel and Conventional CORBA Interface to Java

• 7.2.1ind)CORBA Model:
• 7.2.2ind) Current Activities:
• 7.2.3ind) HPJava Issues:
  

7.3 Task Parallelism and Java for both Integration(Coarse grain) and fine grain examples

Fortran-M CC++, "channel" classes in Java

Relation to MetaChaos and PCRC Interoperability

Maryland will give initial words -- see sections 7.4.2umd and 7.4.3umd

• 7.3.1)Limited Synchronization Capabilities of current Java:

7.4 HPCC and Compute-Webs (Use of World Wide Web MetaComputer)

This is currently a collection of exemplar systems/ideas:

• 7.4.1umd)Mobile Programs:
• 7.4.2umd) Coupling sequential Java Programs:
• 7.4.3umd) Coupling HPJ programs with one another and with other data parallel programs (e.g. MPI, HPF, HPC++):
• 7.4.4syr) WebWork: Integrated Programming Environment Tools for National and Grand Challenges
• 7.4.5syr) Computational Web Multiserver - a proposal for Web based HPCC
• 7.4.6ind) Applet view of HPCC Applications:

7.5 HPCC Java and implications for security

Brokers as interfaces to Java native HPCC related classes

Gannon initial words: see the CORBA discussion in Section 7.2

8 Low Level Runtime Issues

8.1 The Current PCRC Runtime (Regular and Irregular) Implementation in Java

• 8.1.1syr) Introduction:
• 8.1.2syr) Features for PCRC Runtime in Java:
• 8.1.3syr) Issues for PCRC Runtime in Java:

8.2 Distributed Memory Versions of Java Runtime and use of HPCC thread packages on SMP or distributed memory to support Java

• 8.2.1roc) How to implement the thread packages on which Java is based across a range of architectures
• 8.2.2roc) How to unify shared-memory and message-passing portions of web-spanning applications.
• 8.2.3ind) Performance of Current Java Threads:
• 8.2.4ind) Java Distributed Memory Model:
• 8.2.5ind) Implications of Current Indiana Research on HPC++ and CORBA:

back Full Collection of Initial Text

back 1. What are Suitable Models for Parallel Computing and Java

Texas Contribution (28 feb 96):

• 1.1tex) The Nature of Java Applications
• 1.2tex) Consistency with Current Language Philosophy and Feature Set.
• 1.3tex) Inferences
• 1.4tex) Current Features for Specification of Parallel Execution Behavior
• 1.5tex) A Proposal for Additional Specification of Parallelism for Java
• 1.6tex) What Should NOT be Done

  Indiana Contributions 24 feb
  

• 1.4ind)Current Features of Java
• 1.5ind)Extended Synchronization in Java:
• 1.6ind) Is Data Parallelism relevant for Java:
• 1.7ind)Current Strengths and Weaknesses of Java:

back 2. Motivating Applications for HPCCJava

Syracuse Contribution (23 feb 96):

2.0 Motivating Applications for HPJava: Meta-Challenges

2.1Syr) Introduction:

2.2Syr) Meta-Challenges:

2.3Syr) An Example Meta-Challenge: DIS:

2.4Syr) Java for DIS:

Maryland Contribution(22 feb 96):
2.5umd) Remote sensing data fusion:

2.6umd) Interventional Epidemiology/ Medical crisis management:

back 3. Lessons from and Relations to HPF/HPC++

See also Maryland section 7.4.3

Indiana Contributions 24 feb

• 3.1ind) Comparison with HPF: 3.2ind) Comparison with HPC++: 3.3ind) Operator Overloading: back 4. Language Enhancements(Annotations) and Classes to support HPCC and Java
  

Cooperating Systems Contribution (29 feb 96)

Language extensions to Java can be broadly put into two categories, each with its own pros and cons.

• 4.1csc) Extensions via class libraries.
• 4.2csc) Extensions to the Java language.
• 4.3csc) Implications of Immature Language:
• 4.4csc) Comparison with C++: Indiana Contribution (24 feb 96)
• 4.2ind)Implications of Java Philosophy:
• 4.5ind)High Performance Classes:

back 5. Architecture (Shared, Distributed Shared, and Distributed) Issues for HPCC Java

Rochester Contribution(28 Feb 96):

• 5.1roc) Generate native code for shared-memory architectures:
• 5.2roc) Java vs Bytecode:
• 5.3roc) Locality optimizations for (distributed) share-memory architectures:
• 5.4roc) How to integrate shared-memory environments across a heterogeneous collection of machines and interconnects, respecting the dramatically different latencies and bandwidths:
• 5.5roc) Compiling Java for heterogeneous networks of machines:
  Indiana Contribution (24 feb 96):
• 5.6ind)General Issues:

back 6. Compilers and Java -- Both conventional and HPCC and on client(interpreter) and Server(javac)

Rice Contribution(27 feb 96):

• 6.1rice) Compiled vs. Interpreted
• 6.2rice) Compilation issues
• 6.3rice) Interaction with Runtime
• 6.4rice) Tools Interface

  
  Indiana Contribution (24 feb 96):

• 6.1ind) Performance Issues:

back 7. High Level Runtime Issues

back 7.1 Dynamic Communication and Distributed Nameservers

Texas Contribution (29 feb 96):

• 7.1.1tex) Communication Requirements of Java Applications
• 7.1.2tex) The Current Model of Communication
• 7.1.3tex) Desirable Properties of Model of Communication for Java
• 7.1.4tex) Proposed Model of Communication
• 7.1.5tex) Integration/ Implementation Concepts
• 7.1.6tex) Comments

back 7.2 Parallel and Conventional CORBA Interface to Java

Indiana Contribution (24 feb 96):

• 7.2.1ind)CORBA Model:
• 7.2.2ind) Current Activities:
• 7.2.3ind) HPJava Issues:

back 7.3 Task Parallelism and Java for both Integration(Coarse grain) and fine grain examples

see sections 7.4.2umd and 7.4.3umd

Indiana Contribution (24 feb 96):

7.3.1)Limited Synchronization Capabilities of current Java:

back 7.4 HPCC and Compute-Webs (Use of World Wide Web MetaComputer)

There are several important possibilities here -- exemplars follow:

Maryland Contribution(22 feb 96):
7.4.1umd)Mobile Programs:

7.4.2umd) Coupling sequential Java Programs:

7.4.3umd) Coupling HPJ programs with one another and with other data parallel programs (e.g. MPI, HPF, HPC++):

Syracuse Contribution(11 feb 96):

7.4.4syr) WebWork: Integrated Programming Environment Tools for National and Grand Challenges

7.4.5syr) Computational Web Multiserver - a proposal for Web based HPCC

Indiana Comment (24 feb 96):

• 7.4.6ind) Applet view of HPCC Applications:

back 7.5 HPCC Java and implications for security

see the CORBA discussion in Section 7.2

back 8 Low Level Runtime Issues

back 8.1 The Current PCRC Runtime (Regular and Irregular) Implementation in Java

Syracuse Contribution(23 Feb 96):
8.1.1syr) Introduction:

8.1.2syr) Features for PCRC Runtime in Java:

8.1.3syr) Issues for PCRC Runtime in Java:

back 8.2 Distributed Memory Versions of Java Runtime and use of HPCC thread packages on SMP or distributed memory to support Java

Rochester Contribution (28 feb 96):

• 8.2.1roc) How to implement the thread packages on which Java is based across a range of architectures
• 8.2.2roc) How to unify shared-memory and message-passing portions of web-spanning applications.
  

Indiana Contribution24 feb 96):

• 8.2.3ind) Performance of Current Java Threads:
• 8.2.4ind) Java Distributed Memory Model:
• 8.2.5ind) Implications of Current Indiana Research on HPC++ and CORBA: