Full HTML for

Basic foilset Computational Science and HPCC Infrastructure for GEM

Given by Geoffrey C. Fox at GEM Group Meeting Boulder Colorado on March 13-14 1999. Foils prepared March 26 99
Outside Index Summary of Material

We describe a general process
Requirements
Needed Infrastructure
This divides computational effort into 7 areas which are briefly discussed
Describe features of a web-based toolkit

Table of Contents for full HTML of Computational Science and HPCC Infrastructure for GEM

 Denote Foils where Image Critical
Denote Foils where HTML is sufficient
1 Computational Science and HPCC Infrastructure for GEM General Earthquake Simulation Project GEM Meeting March 13,14 99 Boulder
2 Abstract of Computational Science and HPCC Issues for GEM
3 Comments on GEMCI Process
4 Draft Requirements of GEMCI
5 Features of GEM
6 Components of GEMCI: GEM Computational Infrastructure
7 Details of GEMCI - Core Infrastructure I
8 Details of GEMCI - Core Infrastructure II
9 Details of GEMCI - I
10 WebFlow over Globus for NCSA Alliance Quantum Chemistry Application View
11 WebFlow on Globus -- LMS at CEWES
12 Example of a custom Web User Interface Land Management System
13 Details of GEMCI - II
14 Details of GEMCI - III
15 XML for GEM ?
16 7: Overall Integration of GEMCI into a PSE(Problem Solving Environment)
17 Basic 3 Tier Computing Model
18 Object View of running a program
19 GEMCI Architecture
20 Looking into more detail

Outside Index Summary of Material

HTML version of Basic Foils prepared March 26 99

Foil 1 Computational Science and HPCC Infrastructure for GEM General Earthquake Simulation Project GEM Meeting March 13,14 99 Boulder

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Geoffrey Fox
Syracuse University
NPAC
111 College Place Syracuse NY 13244 4100
3154432163
HTML version of Basic Foils prepared March 26 99

Foil 2 Abstract of Computational Science and HPCC Issues for GEM

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
We describe a general process
Requirements
Needed Infrastructure
This divides computational effort into 7 areas which are briefly discussed
Describe features of a web-based toolkit
HTML version of Basic Foils prepared March 26 99

Foil 3 Comments on GEMCI Process

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
GEMCI = GEM Computational Infrastructure
Analysis of field to determine requirements of GEMCI a.k.a. Computational Infrastructure for earthquake science
Infrastructure is hardware; simulation software methodology; algorithms (e.g. fast multipoles); communication infrastructure; databases; security/collaboration/visualization services ...
Ongoing set of activities ("projects") which can motivate and test infrastructure
  • Choose projects so that together exercise "all" aspects of infrastructure
  • KDI is one of these projects?
Training (general computation from PACI, geophysical specific from GEM community)
What is GEMCI Scope?
  • Support for KDI, "GEM" or "Earthquake Science" ?
What is GEMCI Initial Implementation? Support for KDI with a more general architecture?
HTML version of Basic Foils prepared March 26 99

Foil 4 Draft Requirements of GEMCI

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Support distributed collaborators sharing programs, experimental, observational and program-generated data
  • data includes sensors, field observation and commercial GIS
Support execution of programs on sequential and parallel machines -- latter include shared memory systems and clusters of PC's
Support (database and ASCII file based) storage and access to data with necessary metadata and organizational principles
Develop web-based interfaces to simulations and data for both GEM community and outside users
Allow International (ACES) collaboration
Support a range of models and model components in a "plug and play" fashion
Support interactive analysis and visualization
Provide online help and training
HTML version of Basic Foils prepared March 26 99

Foil 5 Features of GEM

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Some Observations:
  • a) GEM is an HPCC class problem as expected to need, for initial Green's function models, some 107 elements and compute power of some 1-100 teraflop. Estimate comes from comparing GEM multipole with related astrophysics problems and the measured extrapolation of these.
  • b) GEM is a relatively young field and is not as obliged as other fields to worry about legacy codes. It should be able to aggressively take advantage of the emerging distributed object web technologies.
  • c) GEM has interesting mix of data and novel algorithms
HTML version of Basic Foils prepared March 26 99

Foil 6 Components of GEMCI: GEM Computational Infrastructure

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Core Infrastructure
1: Web-based User Interface
2: "Collaboration specific" simulation modules built on "industry standards"
3: Geophysical specific modules such as local Physics and friction
4: Evaluation, Data analysis and Visualization
5: Data Storage, indexing and access for experimental and computational information --
6: (Java) Complex Systems and Pattern Dynamics Interactive Rapid Prototyping Environment (RPE) for developing new phenomenological models -- RPE includes analysis and visualization aspects
7: Overall Integration of GEMCI into a PSE(Problem Solving Environment)
HTML version of Basic Foils prepared March 26 99

Foil 7 Details of GEMCI - Core Infrastructure I

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Computational infrastructure involves link of geographically distributed observations and computation
  • Seismic sensors, SAR etc.
  • So hardware supports computation, data storage and access
IGNORE Special Purpose Computers (such as GRAPE in Japan) for O(N2) particle dynamics which could for instance be used in Green's function approach to equation solvers.
  • These have 100 times performance of "conventional parallel machines" but maybe there are much larger algorithmic improvements to be gained which require classic parallel computers
Assume GEM simulations will use classic mix of computers from PC's, Workstation. PC clusters, up to Origin/SP MPP's
Choose software (e.g. Fortran/C+MPI) supporting this range
HTML version of Basic Foils prepared March 26 99

Foil 8 Details of GEMCI - Core Infrastructure II

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Will need access to machines for simulations
  • "your own" for "small" simulations
  • NSF/DoE supercomputer centers for "large" simulations
Need databases (Oracle) for data and metadata
  • What volume is needed?
  • How and where should it be stored
Need communication links
  • Internet2/vBNS is improving networking
Collaboration involves "asynchronous" (off-line) and "synchronous" (interactive)
  • Asynchronous includes web pages and electronic mail
  • Synchronous includes chat rooms, whiteboards, audio-video conferencing, shared data display etc.
    • e.g. NPAC's TangoInteractive or NCSA's Habanero
  • Collaborate inside and outside community (crisis managers)
HTML version of Basic Foils prepared March 26 99

Foil 9 Details of GEMCI - I

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
1: User Interface -- Design and build Web Interfaces to control execution of computational and data access modules.
  • Take advantage of ongoing national DATORR (Desktop Access TO Remote Resources) activity which will lead to standard interfaces between clients, middleware and backend machines and data repositories
  • NCSA is starting major activity in this area for other disciplines ("portal" to solid earth science)
  • No magic bullet. Still tiresome to develop new code and rather poor parallel computing environments
  • Web and distributed objects support "distributed components" and their integration into powerful systems with convenient GUI
    • Components need to be quite big to be efficient
HTML version of Basic Foils prepared March 26 99

Foil 10 WebFlow over Globus for NCSA Alliance Quantum Chemistry Application View

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
HTML version of Basic Foils prepared March 26 99

Foil 11 WebFlow on Globus -- LMS at CEWES

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
WebFlow
server
WebFlow
server
WebFlow
server
EDYS
CASC2D
Data Retrieval
High Performance SubSystem
CASC2D
proxy
IIOP
Web Browser
Data Wizard
WMS interface
Toolbar
HTTP
WMS
File Transfer
File Transfer
GLOBUS
Internet
WebFlow modules
(back-end)
WebFlow
middle-tier
WebFlow applet
(front-end)
HTML version of Basic Foils prepared March 26 99

Foil 12 Example of a custom Web User Interface Land Management System

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Navigate and choose an existing application to solve the problem at hand. Import all necessary data.
Retrieve data
Pre/post-processing
Run simulations
Select host
Select model
Set parameters
Run
HTML version of Basic Foils prepared March 26 99

Foil 13 Details of GEMCI - II

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
2: Large Scale Equation Solver: Exploit collaboration with Caltech and Los Alamos which has developed highly efficient multipole solver for large scale parallel machines (including PC cluster). Major HPCC application. Also need cellular automata and other simulation approaches. All of these methods are expected to parallelize well
3: Local Physics and Friction modules. Develop common interfaces to allow easy experimentation with different approaches
What simulation module languages should we support?
  • Fortran Dominant HPCC simulation language
  • C/C++ Several HPCC successes but not clearly taking off
  • Java (?) Problematical Performance if 100% pure ....
  • Safest is Fortran/C plus MPI with Java "wrappers" if necessary
HTML version of Basic Foils prepared March 26 99

Foil 14 Details of GEMCI - III

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
4: Evaluation, Data analysis and Visualization. Take advantage of partner expertise.
  • For instance Boston University, NCSA and NPACI for visualization of large scale computations;
  • Current toolkit activity includes visualization but do you have existing preferred solutions?
  • Are "immersadesks/CAVEs" i.e. high end 3D immersive devices important?
5: Data Storage, indexing and access for experimental and computational information.
  • Here new distributed object approaches seem powerful and both NPACI and DoE have particularly strong programs that we can leverage.
  • XML standards helpful
HTML version of Basic Foils prepared March 26 99

Foil 15 XML for GEM ?

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
XML is the web distributed object model or
  • XML is ASCII format for database export or
  • XML is universal object serialization technology
If not XML, would need to design lots of database schema, "CORBA/Java Interface Definitions ..."
XML is just a generalized HTML (or a simplified SGML)
<faultsegment id="park101" date="Jan 1 2000" author="" source= .. Lat1="" lat2="" long1="" long2="" depth="" ..>Part of the Parkfield System</faultsegment> <stress type ="prediction" fault="park101" source="Multipole Simulation">(2.3,7.7,-7.2)</stress>
There are/will be wonderful tools to produce access edit and display XML
Need to design GEM specific tags
HTML version of Basic Foils prepared March 26 99

Foil 16 7: Overall Integration of GEMCI into a PSE(Problem Solving Environment)

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Develop a web-based tool-kit (tool-kit, portal, PSE similar concepts) for GEM
Use commercial distributed object web approach which is growing in popularity
  • for instance, NPAC has successfully applied this in a few NCSA and DoD "Problem Solving Environments". Such a commodity software system naturally tracks rapid evolution of technologies and preserves rich functionality (UNIX and Windows compatibility, easy access to databases etc.) Leads to more maintainable systems.
As object web is still under development, recommend modest effort compatible with general principles. Details will emerge ....
Include in PSE support for data manipulation and simulation; composition of components
HTML version of Basic Foils prepared March 26 99

Foil 17 Basic 3 Tier Computing Model

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
A server accepts input and produces output
  • A Web Server accepts HTTP request and returns a web page
  • a Database Server accepts a SQL request and returns records selected from database
  • An Object Broker accepts IIOP requests to invoke methods of an "object" (e.g. run a program)
IIOP and HTTP are two common protocols (formats of control data) for inter program messages
A Web browser (Netscape or Microsoft) can access any server at "the click of a button" with data from user refining action
HTML version of Basic Foils prepared March 26 99

Foil 18 Object View of running a program

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Similar to invoking a web page
Object Broker
Fortran Simulation Code on Sequential or
Parallel Machine
Convert Generic Run Request into Specific Request on Chosen Computer
Fortran Program
is an Important
Type of Object
It can be built up from
smaller objects
e.g. Multipole
library could be an
object
HTML version of Basic Foils prepared March 26 99

Foil 19 GEMCI Architecture

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
Application Integration
Visualization Server
Seamless Access
Collaboration
Security Lookup
Registration
Agents/Brokers
Backend Services
Middleware
Bunch of
Web Servers
and Object
Brokers
HTML version of Basic Foils prepared March 26 99

Foil 20 Looking into more detail

From Computational Science and HPCC Infrastructure for GEM GEM Group Meeting Boulder Colorado -- March 13-14 1999. *
Full HTML Index
The HPCC computing resource breaks into more detail
  • Programming Model: MPI, OpenMP
  • Generic Libraries: Multipole and other solvers, Matrices, I/O, visualization ...
  • GEM Specific Libraries: friction ....
  • Hardware architectures and resources
© 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 Sat Mar 27 1999