ACES Working Group 5 Meeting Summary
Computational Environments and Algorithms

Geoffrey Fox (1), David Place(2), Hiroshi Okuda(3), David Yuen(4), Nai-Gang 
Liang(5)
(1) Florida State University, USA. (2) QUAKES, Australia. (3) University of Tokyo, 
Japan. (4) University of Minnesota, USA. (5) CAP, CSB, China.
Abstract
We describe the computational environments and algorithms working 
group 5 of ACES. We discuss, goals, vision, process and some near 
term activities and interests of the working group members. We 
established 5 subgroups and identified initial coordinators.

Introduction
We devoted a productive session on Thursday Oct 19 2000 at the ACES conference in 
Hakone Japan to setting up the computational environments and algorithms working 
group. The result of this meeting is described here. First we proposed as the initial vision:
Identify, promote and establish computational techniques, environments and 
infrastructure to enable international collaboration to advance research and 
education in earthquake science.

Goals of ACES Working Group 5
1) Establish cross working group standard Challenge problems and datasets
2) List, organize and exchange software, algorithms and data
3) Promote the standards and best practice methodology to ensure that software and data 
can be coupled together in multi-disciplinary applications and can run on the major 
machines of the ACES participants
? This includes parallelization and visualization approaches
? and (XML) data structure standards
4) Track technology and report to community
5) Identify ways of reducing unnecessary duplication by enabling re-use and 
encouraging particular projects to address international requirements
6) Establish Geoscience specific requirements of generic information technology/HPCC 
techniques like visualization, computer architecture etc.
7) Establish information resource for articles and links of importance from inside and 
outside community (e.g. XML standards efforts in other fields)
8) Develop education material of Geoscience specific information 
technology/computational science and promote use in universities, conference 
tutorials and public outreach


ACES Working Group 5 Process
1) Set up basic organization in first two months aiming at major successes by time of 
next ACES meeting (18 months). Develop publishable results and significant 
education/web resources in next year.
2) Agree on 5 sub-groups -- each with at least two coordinators spanning ACES 
countries.
3) List initial expertise / relevant projects of WG5 participants
4) Solicit industry participation -- Hitachi. NEC, IBM, SGI, Sun ….
5) Identify joint projects (such as LSMearth/GeoFEM collaboration; benchmark 
GeoFEM on USA machines)
6) Identify funding for WG5 activities

ACES Working Group 5 Subgroups
Here we list subgroups and initial coordinators.
1) Challenge problems: Fox and Okuda
2) Visualization: Yuen and Fujishira
3) Middleware, Coupling and Data Structures: Parker and Lizuka
4) Algorithms and Parallelization: Place and Nakajima
5) Education: Kellogg and another coordinator (preferably from outside USA)


ACES Working Group 5 Near Term Activities
1) Send around information (minutes of first meeting) about WG5 to ACES community 
to gather initial working group members and information about their expertise and 
(local) activities
2) Develop a “request” for other ACES working groups to gather input for WG5 
subgroups – especially focus on challenge problems defined below.
3) Challenge problems e.g large parallel granular motion code (requires load balancing 
and complex strategies which have been developed by other communities)
? Visualization
? Coupling
4) Identify possible near term successes (low hanging fruit) -- for instance
? Benchmark “GeoFEM on ASCI” or “Virtual California on Earth Simulator” 
? Paper on “Visualization in Earth Science” (issues, tools, experience)
5) Establish (distributed) web resource for WG5
? JPL will initially coordinate this
6) Special Issue of Concurrency and Computation: Practice and Experience (Wiley) of 
Conference (and perhaps other) papers aimed at Computer Science community. This 
will be coordinated with the general ACES publication plan and has same deadlines. 
In spite of the arcane title, this journal focuses on practical computer science or 
computational science. Articles whose message is mainly the computational or 
numerical method or issues are suitable. Authors are encouraged to include 
geoscience background and results to provide context and requirements.
1) Submitted papers due end of January 2001
2) Final papers due May 2001
3) Publish September 2001
4) Hope to have strong representation from GeoFEM group but also papers from full 
ACES community


WG5 Participants: Interests and Subgroups
Here we list people, some of their interests and the subgroups with which they could be 
associated. This only lists people who attended the Thursday meeting and others will be 
added.
1) Geoffrey Fox Education, middleware, parallel algorithms; 2 3 4 5
2) David Yuen visualization and algorithms, molecular dynamics, CFD, Education ;2 
and 5
3) David Place QUAKES, LSM/DEM, Visualization, object methods ;1,2,3,4
4) Peter Mora Micro/macro simulation/ACES Ex-officio
5) David Sparks granular physics, DEM, CFD, Education (earthquakes liked by public) 
;5
6) Jay Parker  GPS, quasistatic VE model,assimilation,;1,3,&4
7) Gerald Hofer LSM, parallelisation, visualization ; 2,4
8) Steffen Abe LSM, gouge dynamics, friction laws, Object methods, parallelization ;4
9) Kengo Nakajima GeoFEM, linear solver/adaptation, Preconditioners, boundary 
element and FEM methods ;1,3&4
10) Yoshitaka Wada GeoFEM, mesh generation/pre-processing FEM;3
11) Hiroshi Okuda GeoFEM, Leader, Parallel FEM, CFD, meshless methods, 
middleware enabling interoperability ; 1 3 4 5
12) Kazuro Hirahara GeoFEM earthquake cycle model, propagation ;3
13) Hiroaki Matsui GeoFEM, Earth's core dynamics, MHD	;1 & 4
14) Mikio Iizuka GeoFEM, nonlinear, FEM and Finite difference CFD, coupling 
(LSMearth GeoFEM) ;3 & 4 
15) Li Chen GeoFEM, visualization (parallelization, computer graphics) ; 2
16) Issei Fujishiro GeoFEM, Visualization (architecture, representation of complex 
fields, interactive design) ;2
17) Kazuo Minami GeoFEM, Performance Optimization/Solver ;1
18) Hisashi Nakamura GeoFEM,CES (Computational Earth Science) Director, does 
climate as well ;5
19) Kazuteru Garatani GeoFEM, static/dynamic solid mechanics ;1 & 4	
20) Hideki Matsumoto NEC-RVSLIB, Visualization
21) Osamu Hazama GeoFEM parttime student, exchange with QUAKES on LSMearth-
GeoFEM coupling, meshfree methods, parallelization