NCTS 2006 School on Modern Numerical Methods in Mathematics and Physics : Grid Computing,Taiwan's Perspectives

(1)

Grid Computing,

Taiwan’s Perspectives

Simon C. Lin

ASGC, Academia Sinica

Taiwan

(2)

Induction: What are Grids and e-Science? –May 18th, 2004 -

2

Exponential Growth World

Gilder’s Law (32X in 4 yrs) Storage Law (16X in 4yrs) Moore’s Law (5X in 4yrs)

Triumph of Light – Scientific American. George Stix, January 2001

Performance per Dollar Spent

Optical Fibre

(bits per second)

Chip capacity

(# transistors)

Data Storage

(bits per sq. inch)

Number of Years

0 1 2 3 4 5 9 12 18

Doubling Time

(3)

"#$%#&'()**'+$,-)'./$,0)'/)$1#$23&0)4

!

! !

!

"#$%! ! &''()$+),-!*

! ./0+#1!

2!'#%!34(,'0!

5676'

'

8)9)$:#,$'2#%)*,&;'

'

<=>?'

?@!AA'

566A'

'

B*)0C$#&,0'9C$D0CD$)'

'

E+F<'

5@?!A'

566?'

'

+#*G2)$'%G&32,09'

' 0*D9C)$'

5@AAA'

566H'

'

E23;)'3&3*G9,9'

'

0D9C#2'

5!I'

566I'

'

JD3&C'2#)0D3$'%G&'

' 0*D9C)$'

HHH'

566!'

'

<#2/'1*D,%'%G&32,09'

' 0*D9C)$'

?K7'

566L'

'

B*)0C$#&,0'9C$D0CD$)'

'

F"E'

5!6'

566K'

'

"$3:,C3C,#&'

' 0*D9C)$'

!L'

5667' ' JD3&C'0M$#2#%G&'

'

0D9C#2'

5?N!'

5666'

'

"$3:,C3C,#&'

'

0D9C#2'

LN6'

?AAA'

'

<#2/'1*D,%'%G&32,09'

' 0*D9C)$'

5N6'

?AA5'

'

FC$D0CD$3'3&3G9,9'

' 0*D9C)$'

AN?I'

'

!"#$%"$&'$(%

")%*+,-./#&'%

.-%01%2'+$(%

(4)

4 !"#$$%&'()*+,-.&/0121314,5),67&819167&8*::1 !

"#$%#&'()**'+$,-)'#./012)3'4##$)53'617

!"#$%"$&'$(% ")%*+,-./#&'% .-%01%2'+$( !"#$"%&'""#( !"#$"%&)(** ++,(-.&'""#(// 89:8;<= <>:8?@@@

(5)

4

“… A new age has dawned in

scientific and engineering

research, pushed by

continuing progress in

computing, information, and

communication technology,

and pulled by the expanding

complexity, scope and scale

of today’s challenges. The

capacity of this technology

has crossed thresholds that

now make possible a

comprehensive

cyberinfrastructure on which

to build new types of scientific

and engineering knowledge

environments and

organizations, and to pursue

research in new ways and

with increased efficacy…”

(6)

Water, water, everywhere,

nor any drop to drink

(7)

The Data Deluge

• A large novel:

1 Mbyte;

The Bible:

5 Mbytes

• A Mozart symphony (compressed):

10 Mbytes

• A digital mammogram:

100 Mbytes

• OED on CD:

500 Mbytes

• Digital movie (compressed):

10 Gbytes

• Annual production of refereed journal literature (∼20 k journals;

∼2 M articles):

1 Tbyte

• Library of Congress:

20 Tbytes

• The Internet Archive (10 B pages) (From 1996 to 2002):

100 Tbytes

• Annual production of information (print, film, optical & magnetic

media):

1500 to 3000 Pbytes

(8)

(9)

The LHC Data Challenge

• Starting from this event

• You are looking for this “signature”

Selectivity: 1 in 1013

Like looking for

1 person in a thousand world populations! Or for a needle in 20 million haystacks!

(10)

The Computing Needs will grow

• It may grow to the scale of ExaBytes of

Data and PetaFlops Computing by 2015, in

particular, the luminosity will be enhanced

even in early stage

• The largest commercial database currently

can only handle tens of TeraBytes

• The fastest stand-alone computer now is

only capable of delivering 70 TeraFlops

(11)

Enabling Grids for E-sciencE

Tera → Peta Bytes

• RAM time to move

– 15 minutes

• 1Gb WAN move

time

– 10 hours ($1000)

• Disk Cost

– 7 disks = $5000 (SCSI)

• Disk Power

– 100 Watts

• Disk Weight

– 5.6 Kg

• Disk Footprint

– Inside machine

• RAM time to move

– 2 months

• 1Gb WAN move

time

– 14 months ($1 million)

• Disk Cost

– 6800 Disks + 490 units + 32 racks = $7 million

• Disk Power

– 100 Kilowatts

• Disk Weight

– 33 Tonnes

• Disk Footprint

– 60 m2

(12)

Academia Sinica, Taiwan

INFSO-RI-508833 12

Mohammed & Mountains

• Petabytes of Data cannot be moved

It stays where it is produced or curated

Hospitals, observatories, European Bioinformatics Institute, …

A few caches and a small proportion cached

• Distributed collaborating communities

Expertise in curation, simulation & analysis

• Distributed & diverse data collections

Discovery depends on insights

Unpredictable sophisticated application code

Tested by combining data from many sources Using novel sophisticated models & algorithms

• What can you do?

– If the mountain won't come to Mohammed, Mohammed must go to the mountain

(13)

From Optimizing Architecture to

Optimizing Organisation

• High-performance computing has

moved from being a problem of

optimizing the architecture

of an

individual supercomputer to one of

optimizing the organization

of large

numbers of ordinary computers

operating in parallel.

(14)

14

Simulated Time Can Get Rough

to ~20 attoliters (1 al = 10–18 _{l), and}

near-field scanning optical microscopy (NSOM, panel D) (10), which allows attoliter detec-tion volumes. NSOM is best suited to cell-surface studies where scanning is important. Single-molecule detection in reduced de-tection volumes can also be achieved with stimulated emission depletion (STED, panel E) (11, 12). With this method, detection vol-umes have been reduced to 0.67 attoliters (11). The detection volume can be placed anywhere in solution, including inside a cell. However, use of STED with more than one color of fluorophore is difficult, and the STED beam is of high intensity.

Another possibility is the use of nanoflu-idic channels to restrict the movement of molecules to a width and depth smaller than

the confocal detection volume (panel F) (13). Molecules would not be near metal surfaces, and diffusion times would be longer because diffusion would be effec-tively one-dimensional. Finally, if super-resolution lenses can be made from materi-als with a negative index of refraction, they may in the future be used for single-mole-cule detection in small volumes (14).

All these methods have strengths in dif-ferent areas and are suitable for difdif-ferent ap-plications. The strength of the zero-mode waveguide of Levene et al. (6) is that it al-lows parallel detection of many single mole-cules while providing the smallest detection volumes of any method reported to date, en-abling single-molecule analysis of much weaker interactions than previously possible.

References

1. H. P. Lu, L. Xun, X. S. Xie,Science 282, 1877 (1998). 2. K. Adachi et al., Proc. Natl. Acad. Sci. U.S.A. 97, 7243

(2000).

3. A. A. Denizet al., Proc. Natl. Acad. Sci. U.S.A. 97, 5179 (2000).

4. B. Schuler, E. A. Lipman, W. A. Eaton,Nature 419, 743 (2002).

5. X. W. Zhuanget al., Science 288, 2048 (2000). 6. M. J. Levene et al., Science 299, 682 (2003).

7. C. Zander, J. Enderlein, R. A. Keller, Single Molecule Detection in Solution: Methods and Applications (Wiley-VCH, Berlin, 2002).

8. R. X. Bian, R. C. Dunn, X. S. Xie, P. T. Leung,Phys. Rev. Lett. 75, 4772 (1995).

9. T. E. Starr, N. L. Thompson,Biophys. J. 80, 1575 (2001). 10. F. de Lange et al., J. Cell Sci. 114, 4153 (2001).

11. T. A. Klar et al., Proc. Natl. Acad. Sci. U.S.A. 97, 8206 (2000).

12. M. Dyba, S. W. Hell, Phys. Rev. Lett. 88, 163901 (2002).

13. M. Foquet et al., Anal. Chem. 74, 1415 (2002). 14. J. B. Pendry, Phys. Rev. Lett. 85, 3966 (2000).

H

igh-performance com-puting has moved from being a problem of op-timizing the architecture of an individual supercomputer to one of optimizing the organi-zation of large numbers of or-dinary computers operating in parallel (1). This development has been made possible by rapid progress in microproces-sor, memory, and storage com-ponents.

On page 677 of this issue, Korniss et al. (2) point out a key weakness of this ap-proach. Simulated time, the common parameter linking the many loosely coupled

ele-ments of a distributed simulation, can get rough. Just as the surfaces of crystalline materials grown by depositing individual atoms roughen (3), so does the temporal surface of a computation spread continu-ously as the modeled system evolves.

This roughening becomes a problem when measurements must be continuously extracted from the complex simulation. The solution proposed by Korniss et al. takes ad-vantage of a property found in the Internet: its “small world” nature of having occasional links between points that would otherwise be thought to be far apart. The authors show that intermittent synchronization over random distances can suppress temporal roughening.

The best evidence of the changing na-ture of high-performance computing comes from the regularly updated “TOP500” sur-vey of the world’s 500 biggest computing complexes (4). By 2002, clusters of conven-tional computers had come to make up 93% of the listed machines, with old-fashioned, vector-based supercomputers constituting the remaining 7%. Ten years ago, when the TOP500 surveys began, 27% of the listed systems used a conventional architecture, while 66% were specialized vector mach-ines. The remaining 7% were variations on the massively parallel Connection Machine (5), a now-extinct species.

For some computations, which are vast in extent but simple in organization, a new computing resource with lower costs is be-ing tapped. “Embarrassbe-ingly parallel” computations are increasingly computed at

the “edge of the network.” A computation is distributed to many cooperating worksta-tions, which perform their pieces of the work with little or no coordination as background tasks. They return the completed result to a cen-tral host, which checks the re-sults, assigns new tasks, and compiles the partial results. This approach has been used, for example, for factoring ex-tremely large numbers and for folding a protein. But the best known example is the SETI@home project (6), in which over 4 million users have analyzed data from a ra-dio telescope, seeking evi-dence of narrowband trans-missions at likely communica-tions frequencies. The work is done at the lowest possible pri-ority—it is a screen saver.

The SETI@home project estimates that the total cpu power of the thousands of users active on any given day is 13 teraflops, greater than the combined power of the world’s three biggest supercomputers, as listed in the 2001 TOP500 survey (4). However, the co-ordination involved in distributing and col-lecting all the work for the SETI@home project has consumed a significant fraction of the communications bandwidth to the outside world from the University of California’s Berkeley campus.

Efforts are under way to make approach-es like this applicable to a much wider class of tasks. The GRID consortium (7) of open-source developers is creating an infrastruc-ture that permits jobs to be submitted any-where within a “power grid” of reasonably reliable computer servers, under control mechanisms that will permit costs and re-sources to be allocated fairly. In the GRID

0 100 200 ₃₀₀ 400 500 ₆₀₀ 700 800 ₉₀₀ 1000 0 100200 300400 500600 700800 9001000 490 495 500 505 510 515 520

Time steps completed

Processor

Out of sync. This simulated temporal surface was created by allowing 1000 x 1000 processing elements to proceed in equal time steps under local synchronization.

The author is in the Department of Computer Science and Engineering, Hebrew University, Jerusalem, 91904 Israel. E-mail: kirk@cs.huji.ac.il

P E R S P E C T I V E S

C O M P U T E R S C I E N C E

Rough Times Ahead

Scott Kirkpatrick

(15)

Paradigm Shift in SciComp

• We are about to witness a paradigm

shift in Scientific Computing

• Collaborative Data and Compute

intensive Research (e-Science) will be

everywhere

• Infrastructure Development v.s.

traditional ad hoc HPC system

• What are the challenges to

(16)

It’s about

(17)

Some Quotations

• If I have been able to see further, it was

only because I stood on the shoulders of

giants.

--- Isaac Newton, Letter to Robert Hooke

• The real problem has always been, in my

opinion,

getting people to collaborate on

a solution [for a common problem].

--- David Williams, from

“50 years of Computing at

CERN”

(18)

2nd_{EGEE Conference – Den Haag - 23}rd_{November 2004}

INFSO-RI-508833 18

Why work together

• Wonderful opportunity

–

Can do things that can’t be done alone – its more fun!

–

Recognising and Establishing e-Dreams

–

Combine our creativity, ingenuity and resources

• Challenge so Hard – can’t “go it alone”

–

Building a broad user community is hard

–

Building e-Infrastructure is hard

–

Deploying e-Infrastructure is hard

–

Sustaining e-Infrastructure is very hard

• Realising our potential

–

Multipurpose & Multidiscipline infrastructure

–

Amortise costs

–

International competition and collaboration

(19)

“Rules of Engagement”

• A foundation of Honesty

–

Admit what we don’t understand

–

Admit the limits of our software

–

Admit the limits of our support

• Realism

–

Working together is worthwhile – so is competition – choose!

–

But they both take time to show a “profit” – persist

• Openness and Mutual Respect

–

Stand shoulder-to-shoulder to meet challenges

–

Be prepared to adapt your ideas and plans

• Commitment

–

Make strong / explicit commitments – or say “no”

(20)

Grid - its really about collaboration!

• It’s about sharing

and building a

vision for the

future

– And it’s about getting

connected

• It’s about the

democratization

of science

• It takes advantage

of Open Source!

(21)

Success on a Worldwide scale

• If we can bring together people from all over the

world (whether they be physicists, biologists,

computer scientists, climate researchers or ….) and

they

– Want to be part of building the “cyber infrastructure” or Grid environments or “e-science environments” for the future

– Actively participate

– Get benefit from the collaboration

(22)

Any Opportunity for

Small Countries?

(23)

LCG Service Hierarchy

Tier-0 – the accelerator centre

 Data acquisition & initial processing  _{Long-term data curation}

 _{Distribution of data  Tier-1 centres}

Canada – Triumf (Vancouver) France – IN2P3 (Lyon)

Germany – Forschunszentrum Karlsruhe Italy – CNAF (Bologna)

Netherlands Tier-1 (Amsterdam)

Nordic countries – distributed Tier-1

Spain – PIC (Barcelona)

Taiwan – Academia SInica (Taipei) UK – CLRC (Oxford)

US – FermiLab (Illinois) – Brookhaven (NY)

Tier-1 – “online” to the data

acquisition process  high availability

 Managed Mass Storage –

 grid-enabled data service

 _{Data-heavy analysis}

 National, regional support Tier-2 : ~120 centres (40-50 federations) in ~29 countries  Simulation

(24)

October 7, 2005

LCG Status report Ian.Bird@cern.ch LHCC Open Meeting; 28th_{June 2006}

24

LCG depends on 2 major science grid

The LCG service runs & relies on grid infrastructure

provided by:

EGEE - Enabling Grids for E-Science

(25)

EGEE:

> 180 sites, 40 countries > 24,000 processors,

~ 5 PB storage

EGEE Grid Sites : Q1 2006

0 20 40 60 80 100 120 140 160 180 200 38078 38231 38384 38443 38596 38687 site s 0 5000 10000 15000 20000 25000 30000 38078 38231 38384 38443 38596 38687 38777 CP U EGEE:

(26)

Simon C. Lin NCTS, Hsinchu, Taiwan

OSG Participation



OSG Production: 53 CEs, 18739 CPUs, 9 SEs (Feb. 2005)



4 sites in Asia: ASGC, NTU, and KNU

(27)

Use of the infrastructure

More than 35K jobs/day on the EGEE Grid

LHC VOs  30K jobs/day

Sustained & regular workloads of >35K jobs/day

•

spread across full infrastructure

(28)

INFSO-RI-508833 28

EGEE Operations Process

• Grid operator on duty

– 6 teams working in weekly rotation

 CERN, IN2P3, INFN, UK/I, Ru,Taipei

– Crucial in improving site stability and management

– Expanding to all ROCs in EGEE-II

• Operations coordination

– Weekly operations meetings

– Regular ROC managers meetings

– Series of EGEE Operations Workshops

 Nov 04, May 05, Sep 05, June 06

• Geographically distributed responsibility for operations:

– There is no “central” operation

– Tools are developed/hosted at different sites:

 GOC DB (RAL), SFT (CERN), GStat (Taipei), CIC Portal (Lyon)

• Procedures described in Operations Manual

– Introducing new sites

– Site downtime scheduling

– Suspending a site

– Escalation procedures

– etc

Highlights:

• Distributed operation

• Evolving and maturing procedures

• Procedures being in introduced into and shared with the related infrastructure projects

(29)

Taiwan’s Thinking



International collaboration/ participation with the

global Grid Cyber- and e-Infrastructure is the key



Our strategy is to

use High Energy Physics to drive

the

Next Generation Grid Infrastructure. No national Grid

organization has been established yet !



We are setting up a new Academia Sinica Grid

Computing Centre (ASGC) reorganised from ASCC



ASGC will provide Grid-based infrastructure,

services, e-Science application development, and

promote Grid activities in Taiwan



ASGC is fully supported by Academia Sinica and

(30)

Taiwan signed the first WLCG MoU

• Jos Engelen (CERN CSO) and President YT

Lee (Academia Sinica) signed the WLCG MoU

on 9 Dec. 2005 as a Tier-1 Centre for ATLAS

and CMS

(31)

VN TH CN MY JP KR HK PH TW SG

(32)

Plan for Taiwan Tier-1 Network

Backup Path to T0

Primary Path to T0

(plan to install 10GE in 2006)

(33)

The EGEE-II project

• EGEE

–

1 April 2004 – 31 March 2006

–

71 partners in 27 countries, federated in regional Grids

• EGEE-II

–

1 April 2006 – 31 March 2008

–

91 partners in 32 countries

–

13 Federations

• Objectives

–

Large-scale, production-quality

infrastructure for e-Science

–

Attracting new resources and

users from industry as well as

science

–

Maintain and further improve

“gLite” Grid middleware

(34)

EU-SEA Singapore 2006

EGEE-II INFSO-RI-031688 ₃₄

EGEE – What do we deliver?

• Infrastructure operation

– Currently includes >200 sites across 39 countries

– Continuous monitoring of grid services & automated site configuration/management

• Middleware

– Production quality middleware distributed under business friendly open source licence

• User Support

- Managed process from first contact

through to production usage

– Training

– Documentation

– Expertise in grid-enabling applications

– Online helpdesk

– Networking events (User Forum, Conferences etc.)

• Interoperability

(35)

EGEE-II Expertise & Resources

• 32 countries

• 13 federations



Major and national

Grid projects in

Europe, USA, Asia

+

27 countries

through related projects:

– BalticGrid – EELA – EUChinaGrid – EUIndiaGrid – EUMedGrid – SEE-GRID

(36)

EGEE-II INFSO-RI-031688

EGEE Asia-Pacific Federation

PAEC NCP Islamabad IHEP Beijing KNU Daegu Univ. Melbourne GOG Singapore KEK Tsukuba ICEPP Tokyo

Taipei - ASGC, IPAS NTU, NCU

VECC Kolkata

Tata Inst. Mumbai

• 15 EGEE grid sites in Asia Pacific

• Academia Sinica Grid Computing

Centre (ASGC) is acting as the

coordinator in Asia

• Operation Centre in Asia Pacific Region

• LCG Tier-1 Centre for Physics

• Potential new sites from Thailand,

Malaysia, Australia, New Zealand

(37)

Contributions of ASGC in WLCG

• WLCG Tier1 Centre -- Collaborating ATLAS & CMS

Teams (NCU, NTU, IPAS) in Taiwan

• Regional Operation Centre and Core Infrastructure

Centre

• Production CA Services

• LCG Technology Development

– Data Management

– Grid Technology

– Certification & Testing

CIC

and

ROC

in EGEE

 APROC established in April 2005

 Supports EGEE sites in Asia Pacific

 Australia, Japan, India, Korea,

Singapore, Taiwan



8 sites,

6 countries

 Provides Global and Region services

(42)

Education and Training

Note: gLite and the development of EGEE were introduced in all the events which are run by ASGC

Event Date Attendant Venue

China Grid LCG Training _{16-18 May 2004} 40 Beijing, China

ISGC 2004 Tutorial _{26 July 2004} 50 AS, Taiwan

Grid Workshop _{16-18 Aug. 2004} 50 Shang-Dong, China

NTHU _{22-23 Dec. 2004} 110 Shin-Chu, Taiwan

NCKU _{9-10 Mar. 2005} 80 Tainan, Taiwan

ISGC 2005 Tutorial _{25 Apr. 2005} 80 AS, Taiwan

Tung-Hai Univ. _{June 2005} 100 Tai-chung, Taiwan

EGEE Workshop _{Aug. 2005} 80 20th APAN, Taiwan

EGEE Administrator Workshop Mar. 2006 40 AS, Taiwan

(43)

(44)

 Formation of informal Asia Collaboration Board on 1 December 2005

during LCG Service Challenge meeting in Taipei

 Attendees: Australia, China, Japan, Korea, Singapore, Taiwan

 Mission

 Enable researchers to leverage Grid technologies

 Foster collaboration in Asia and the rest of the world

 Virtual Organization

 APDG: Asia-Pacific Data Grid

 VO for deployment testing and training

 APESCI: Asia-Pacific e-Science

 Production VO to incubate early Grid communities

(45)

Plan of AP Federation



VO Services

:

deployed from April 2005 in Taiwan (APROC)

 LCG: ATLAS, CMS

 BioInformatics, BioMed

 Geant4

 APeSci :

for collaboration general e-Science services in Asia Pacific Areas

 APDG: for testing and testbed only

 TWGRID:

established for local services in Taiwan

 Potential Applications

 LCG, Belle,

nano, biomed, digital archive, earthquake,

– Astrophysics

 MAGIC, Planck

– Computational Chemistry

– Earth Sciences

 Earth Observation, Solid Earth Physics, Hydrology, Climate

– Financial Simulation

 E-GRID

– Fusion

– Geophysics

 EGEODE

– High Energy Physics

 4 LHC experiments (ALICE, ATLAS, CMS, LHCb)

 BaBar, CDF, DØ, ZEUS

– Life Sciences

 Bioinformatics (Drug Discovery, GPS@, Xmipp_MLrefine, etc.)

 Medical imaging (GATE, CDSS, gPTM3D, SiMRI 3D, etc.)

(48)

24 Jan. 2006 21st APAN, Japan

(49)

• Biomedical goal

–

accelerating the discovery of novel potent

inhibitors thru minimizing non-productive

trial-and-error approaches

–

improving the efficiency of high throughput

screening

• Grid goal

–

massive throughput: reproducing a

grid-enabled in silico process (exercised in DC I)

with a shorter time of preparation

–

interactive feedback: evaluating an alternative

light-weight grid application framework

(DIANE)

EGEE Biomed DC II – Large Scale

Virtual Screening of Drug Design

Credit: Y-T Wu N1 H5 Credit: Y-T Wu

(50)

EGEE-II INFSO-RI-031688 ₅₀

Example: Avian flu (I)

• Avian Flu H5N1

– H5 and N1 = proteins on virus surface

(Hemagglutinins and Neuraminidases)

– Neuraminidases play major role

in virus multiplication

• Present drugs (e.g. Tamiflu)

– inhibit action of neuraminidases and stop virus proliferation

• Problems

– Virus keeps mutating

– Drug-resistant N1 variants can appear

• Biological goal of data challenge

– Study in silico the impact of selected point mutations on the efficiency of existing drugs

– Find new potential drugs

N

H

Credit: Y-T Wu

(51)

(52)

mpiBLAST-g2

ASGC, Taiwan and PRAGMA http://bits.sinica.edu.tw/mpiBlast/index_en.php

 A GT2-enabled parallel BLAST runs on Grid  GT2 GASSCOPY API

 MPICH-g2

 The enhancement from mpiBLAST by ASGC

• Performing cross cluster scheme of job execution

• Performing remote database sharing • Help Tools for

– database replication

– automatic resource specification and job submission (with static resource table) – multi-query job splitting and result merging

• Close link with mpiBLAST development team

– The new patches of mpiBLAST can be quickly applied in mpiBLAST-g2

(53)

(54)

24 Jan. 2006 21st APAN, Japan

(55)

(56)

Long-Term Archives for AS NDAP Contents

Project Totel Files Total Size (MB)

珍藏歷史文物 3,353 4,495,853.22 管理員 1,095 981.33 台灣貝類相 3,878 21,869.78 近代中國歷史地圖與遙測影像資訊典藏計畫 33,671 364,554.69 語言典藏計畫 1 7.05 技術研發分項計畫 39,315 98,246.45 魚類資料庫 32,070 4,199.32 台史所 34,040 44,157.20 台灣本土植物 31,027 1,578,654.76 近代外交經濟重要檔案計畫 603,997 20,601,428.38 台灣原住民 601,715 1,516,242.05 1,384,162 28,726,194.23

Table I. Size of Digital Contents of NDAP

2002 2003 2004 2005 Total

Total Data Size (GB) 22,810.00 38,550.00 63,480.00 70,216.02 195,056.02

AS Production (GB) 22,800.68 31,622.17 47,430.79 55,757.47 157,611.11

Table II. Details of NDAP Production in 2005

Metadata Size(MB) Metadata Records Data Size(GB)

All Inst. 56,204.40 1,035,538.00 70,216.02

(57)

(58)

Linux

Atmosphere Databank Architecture

NCU/databank NTNU/dms MCAT / srb001 Users ASCC/srb002 ASCC/lcg00104(TB) ASCC/lcg00105(TB) NTU/monsoon(TB)

NTU/dbar_rs1, dbar_rs_2 ASCC/gis212(TB)_{ASCC/gis252(TB)}

windows

(59)

• Use LAS (Live Access Server) to access the dataset from the

SRB System, and integrate with Google Earth

(60)

Taiwan GeoGrid



Applications



Grid for Geoscience, Earth Science and Environmental Research and



WebGIS/Goole Earth based UI

(61)

(62)

Industrial Program



NSC-Quanta Collaboration



To help Quanta Blade System have best performance for

HPC and Grid Computing



Quanta is the largest Notebook manufacturer in the world



Participants: AS, NTU, NCTS, NTHU, NCHC



Scientific Research Disciplines: Material Science,

Nano-Technology, Computational Chemistry, Bioinformatics,

Engineering, etc.

(63)

Conclusion



Critical mass decides which Grid technology/system

to prevail; Collaboration, Data and Complexity

Reduction are the main themes



We are about to witness Data Deluge in all disciplines

of e-Sciences



Unprecedented way to collaborate on day-to-day basis

will change the sociology of academia life, eco-system

of business world and eventually every one in the

society



NCTS 2006 School on Modern Numerical Methods in Mathematics and Physics : Grid Computing,Taiwan's Perspectives

Grid Computing,

Taiwan’s Perspectives

Simon C. Lin

ASGC, Academia Sinica

Taiwan

Exponential Growth World

"#$%#&'()**'+$,-)'./$,0)'/)$1#$23&0)4

!

! !

! !

!

"#$%! ! &''()*$+),-!

! ./0+#1!

2!'#%!34(,'0!

5676'

'

8)9)$:#,$'2#%)*,&;'

'

<=>?'

?@!AA'

566A'

'

B*)0C$#&,0'9C$D0CD$)'

'

E+F<'

5@?!A'

566?'

'

+#*G2)$'%G&32,09'

' 0*D9C)$'

5@AAA'

566H'

'

E23;)'3&3*G9,9'

'

0D9C#2'

5!I'

566I'

'

JD3&C'2#*)0D*3$'%G&'

' 0*D9C)$'

HHH'

566!'

'

<#2/'1*D,%'%G&32,09'

' 0*D9C)$'

?K7'

566L'

'

B*)0C$#&,0'9C$D0CD$)'

'

F"E'

5!6'

566K'

'

"$3:,C3C,#&'

' 0*D9C)$'

!L'

5667' ' JD3&C'0M$#2#%G&'

'

0D9C#2'

5?N!'

5666'

'

"$3:,C3C,#&'

'

0D9C#2'

LN6'

?AAA'

'

<#2/'1*D,%'%G&32,09'

' 0*D9C)$'

5N6'

?AA5'

'

FC$D0CD$3*'3&3*G9,9'

' 0*D9C)$'

AN?I'

!"#$%"$&'$(%

"#$%! ! &''()$+),-!*

JD3&C'2#)0D3$'%G&'

FC$D0CD$3'3&3G9,9'