Full HTML for

Basic foilset Computational Grids and Commodity Technologies for CFD

Given by Geoffrey C. Fox at AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington on 10-11 November 97. Foils prepared 1 dec 97
Outside Index Summary of Material

Application talks by Manny Salas and Ray Cosner were in counterpoint to two software talks
  • This is second software talk
A computational grid is a large-scale high performance distributed computing system that presents new challenges from the heterogeneity of the nodes and the variable and sometimes poor performance of network latency and bandwidth.
Application support should include both programming the typically data parallel simulation modules and the composition of modules to build complete processing systems.
So the challenges are supporting fine grain concurrency within a single processor, data parallelism within a tightly coupled system and dataflow or other compositional model between nodes of the grid.
Kennedy's talk reviewed the available technologies: auto-parallelism, explicit communication, distributed shared memory, data parallel languages, task or module parallelism, libraries typically supporting SPMD programming, programming tools, resource management, latency tolerance and run-time compilation.
  • Each approach has clear positive and negative features and any successful programming environment must combine many of these together.
  • Any good approach is likely to need performance estimation, whole-program compilation, new approaches to compilation for scripted and visual environments, run-time compilation, libraries and programming support tools.
This presentation by Fox stressed that there were important systems engineering advantages in building computational grid support on "commodity technologies" such as CORBA and the Web.
This allowed one to exploit database, collaboration and the natural software integration capabilities of these impressive commercial systems designed for pervasive information systems such as worldwide commerce and corporate enterprise Intranets.
There are some critical research issues in how best to incorporate high performance into these systems without losing the power and flexibility of the commodity approach.

Table of Contents for full HTML of Computational Grids and Commodity Technologies for CFD

 Denote Foils where Image Critical
Denote Foils where HTML is sufficient
1 Remarks on Applications and Commodity(Web) Technologies for CFD Session
2 Structure(Architecture) of Applications - I
3 Structure(Architecture) of Applications - II
4 Some Classes of Applications
5 Computational Grid and the Web
6 Commodity Web-Server based Metacomputer Capabilities at 3 levels
7 Technology Convergence Roadmap - Overview for Forces Modeling, Integrated Modeling and Testing for DoD
8 General Commodity Service Integration Model
9 Example of WebFlow = AVS/Khoros using Web
10 Web Collaboration Model
11 Collaborative Applications with their participants
12 Use of Web Collaboration Systems in Multidisciplinary Applications
13 Minimal Web based Multidisciplinary Application
14 Comparison of Communication/Linkage Models

Outside Index Summary of Material

HTML version of Basic Foils prepared 1 dec 97

Foil 1 Remarks on Applications and Commodity(Web) Technologies for CFD Session

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
Geoffrey Fox
Syracuse University
NPAC
111 College Place Syracuse NY 13244 4100
3154432163
HTML version of Basic Foils prepared 1 dec 97

Foil 2 Structure(Architecture) of Applications - I

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
Applications are metaproblems with a mix of module and data parallelism
Modules are decomposed into parts (data parallelism) and composed hierarchically into full applications.They can be the
  • Many thousands of separate programs (e.g. structures,CFD ..) used in design of aircraft
  • the various filters used in Khoros based image processing system
  • the ocean-atmosphere components in CASA project
  • The data-base or file system access of a data-intensive application
  • the objects in a distributed Forces Modeling Event Driven Simulation
HTML version of Basic Foils prepared 1 dec 97

Foil 3 Structure(Architecture) of Applications - II

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
Modules are "natural" message-parallel components of problem and tend to have less stringent latency and bandwidth requirements than those needed to link data-parallel components
  • modules are what HPF needs task parallelism for
  • Often modules are naturally distributed whereas parts of data parallel decomposition may need to be kept on tightly coupled MPP
Assume that primary goal of computational grid is to add to existing parallel computing environments, a higher level supporting module parallelism
  • Now if one takes a large CFD problem and divides into a few components, those "coarse grain data-parallel components" can be supported by computational grid technology
  • so no clean division but coarse grain modules are general goal!
HTML version of Basic Foils prepared 1 dec 97

Foil 4 Some Classes of Applications

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
High end (today) to high end(tomorrow)
  • e.g. Astrophysics QCD .....
High end(today) to low end(tomorrow)
  • Today's supercomputer is tomorrow's PC
  • e.g. TeleImmersive Environments
Low end(today) to low end(tomorrow)
  • e.g. commerce, consumer market, education
Low end(today) to very low end (tomorrow)
  • e.g. email
The new application(s) we can't predict
1997
2007
HTML version of Basic Foils prepared 1 dec 97

Foil 5 Computational Grid and the Web

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
The transportation system was built using lessons from and feed up/down from Sports cars, Cadillacs, Model T's, Ford Escorts etc.
Computational Grid will be shaped by and shape all 5 classes of applications on previous foil
  • Everybody would perhaps agree on general statement but perhaps disagree on importance of it
A highish end computational grid will in some sense (to be disagreed on) be influenced by and influence the "Web" which is here defined as "mass-market"/business IntraNet (low to low) use of Internet/distributed Information Systems
HTML version of Basic Foils prepared 1 dec 97

Foil 6 Commodity Web-Server based Metacomputer Capabilities at 3 levels

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
1:User View: Interoperable Web Interface accessing services through Java Compute Services Framework
2:Network of Java Servers provide distributed services with databases, compute engines, collaboratories, object brokers, instruments
  • All control at this level
  • Some coarse grain computing
Back end "Number Crunchers" linked either by communication at level 2 (slowish but easy) or at level 3 (high performance but more work)
Compute processes linked either to servers or together by MPI if parallel
Java Servers
HTML version of Basic Foils prepared 1 dec 97

Foil 7 Technology Convergence Roadmap - Overview for Forces Modeling, Integrated Modeling and Testing for DoD

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
HTML version of Basic Foils prepared 1 dec 97

Foil 8 General Commodity Service Integration Model

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
This metacomputing model can be combined with emerging Web Collaboration, Object broker and database capabilities to support multidisciplinary applications and computational steering
HTML version of Basic Foils prepared 1 dec 97

Foil 9 Example of WebFlow = AVS/Khoros using Web

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
Simulation
Basic Display
Image Filter
is another
module
Output Display after Filter
Runs as a
parallel
module
using
Java Server
host
HTML version of Basic Foils prepared 1 dec 97

Foil 10 Web Collaboration Model

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
Systems such as Habanero or TANGO supports client side electronic societies of people, instruments and their applications.
A network of Java Servers manage sessions and interfaces them to the world wide Server Infrastructure that underlies emerging model of Commodity applications
In CORBA, Web Collaboration systems become a facility to support sharing and replication of objects.
HTML version of Basic Foils prepared 1 dec 97

Foil 11 Collaborative Applications with their participants

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
Command and Control or more generally distributed (tactical) real-time decision support -- decision makers and gatherers of information
Forums and Chattering on the Web -- the world!
Education -- teachers and students (and administrators)
(Tele)medicine -- Doctors (primary and specialist referrals), Nurses, Patients and administrators
Business Enterprise (strategic) Support as in Lotus Notes -- Employees of Business including especially managers
Multidisciplinary Applications including Collaborative design -- teams of multi organization engineers
Computational Steering -- one or more computational scientist and visualization device
HTML version of Basic Foils prepared 1 dec 97

Foil 12 Use of Web Collaboration Systems in Multidisciplinary Applications

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
Systems such as TANGO link people and shared applications such as chat board, audio video conferencing, visualizations, shared white board, common AUTOCAD design and related tools
CFD
TANGO Server
Database
Object Broker
MPP
Structures
MPP
Engineer
+ core
services
Visualization e.g.CAVE
Shared AutoCAD
Engineer
+ core
services
HTML version of Basic Foils prepared 1 dec 97

Foil 13 Minimal Web based Multidisciplinary Application

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
This combines TANGO for collaboration with WebFlow to link server side applications
If necessary WebFlow would support high performance inter-module communication at the backend but it would always implement control and this allows TANGO integration with server side computation
  • So we suggest change in usual HPCC integration model to move control to server level so that coupled applications can be directly integrated with services such as TANGO, databases ....
WebFlow communication model is a dynamic dataflow
Of course other server side compute models are possible and in general need (web-linked) data bases, file systems, object brokers etc.,
HTML version of Basic Foils prepared 1 dec 97

Foil 14 Comparison of Communication/Linkage Models

From Computational Grids and Commodity Technologies for CFD AirFrame Simulation Panel ARPA Sys.Env. Meeting Washington -- 10-11 November 97. *
Full HTML Index
WebFlow supports dataflow model where user must supply routines to process input of data that drives module and output of data for other modules
TANGO supports shared state and user supplies routines that read or write either
  • Total state of application or
  • Change in state of application
Can be done for applications like AUTOCAD as vendor supplies necessary API
CFD
Structures
© Northeast Parallel Architectures Center, Syracuse University, npac@npac.syr.edu

If you have any comments about this server, send e-mail to webmaster@npac.syr.edu.

Page produced by wwwfoil on Mon Dec 1 1997