• Tidak ada hasil yang ditemukan

Komputasi Kinerja Tinggi

N/A
N/A
Protected

Academic year: 2017

Membagikan "Komputasi Kinerja Tinggi"

Copied!
105
0
0

Teks penuh

(1)
(2)

Lembaga Ilmu Pengetahuan Indonesia

Pusat Penelitian Informatika

Tahun 1965 : Lembaga Elektroteknika Nasional

(LEN)

Keppres No. 1 Tahun 1986 : Puslitbang Telkoma, Inkom, Telimek

dan UPT Pusat LEN

Tahun 1990 : UPT LEN diserahkan ke

BPIS (spin-off)

(3)

59% 11%

27%

Fungsional Peneliti&Kandidat, Non Peneliti dan fungsional Umum

Peneliti: 54 Orang

Fungsional Non Peneliti: 3 Orang Penata Teknis: 10 Orang Wan ita 30% Pria 70%

Profil SDM

3% 11%

Fungsional Umum adm: 24 Orang

26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65 Series1 11 15 19 19 11 8 5 3

(4)
(5)
(6)

Text Data model for Weather Data

DNA QR codes of (a). JX426135; (B) JN245997; (c) JN245994; (d) JN632605

Beberapa Hasil Penelitian

Text Data model for Weather Data JN245994; (d) JN632605

(7)

Adaptasi model iklim wilayah Indonesia menggunakan REGCM 4.0

Pemanfaatan RegCM4 (Regional Climate Model) untuk simulasi iklim spesifik untuk wilayah Indonesia.

Perambatan Energi Gelombang

Bertujuan untuk melacak perambatan gelombang dengan potensi energi

Simulasi Curah Hujan di Indonesia

Density plot tinggi dan energi gelombang

Simulasi Dinamika Populasi Nyamuk dengan Cellular

Model dan simulasi dinamika populasi nyamuk merupakan studi bidang komputasi biologi untuk memahami perubahan ukuran populasi suatu spesies. Hasil simulasi menunjukkan model yang diusulkan mampu mensimulasikan populasi nyamuk secara temporal dan spasial.

Bertujuan untuk melacak perambatan gelombang dengan potensi energi yang cukup besar. Penelitian berfokus pada perambatan energi gelombang permukaan air.

Simulasi Dispersal Nyamuk

(8)

Pengembangan Algoritma Sistem Uji Berbasis Visual

Sistem Uji Berbagai Parameter Kualitas

Pengujian Cip Sensor

Produksi massal CERN, Swiss

Implementasi Pengujian pada Produksi Massal

2015

2016

Pengujian Visual

Cip hasil produksi Cip

(9)

Riset @P2Informatika

(10)

Jenis Layanan

Layanan Komputasi untuk Publik

(11)
(12)

Fasilitas

Cibinong Bandung

Gedung Pusat Inovasi

Jl. Raya Jakarta-Bogor KM 47

Cibinong, Jawa Barat

Gedung 10 Kompleks LIPI

Jl. Cisitu No. 21

(13)

Fasilitas HPC

Master Node (4 Node)

 Prosesor: 2 x 8 core

Intel Xeon E5 Family

 Memori: 128 GB

 Storage: 24 TB

Basic Node (114 Node)

GPU Node (20 Node) - Prosesor: 2 x 4 core

Intel Xeon E5 Family - Memori: 8 - 16 GB - Storage: 500 GB

- GPU Tesla M2075 (488 core)

High Memory Node (8

Basic Node (114 Node)

 Prosesor: 2 x 4 core

Intel Xeon E5 Family

 Memori: 8 - 16 GB  Storage: 500 GB

High Memory Node (8 Node)

- Prosesor: 2 x 8 core Intel Xeon E5 Family - Memori: 256 GB

(14)

HPC LIPI @P2Informatika

Cibinong

928 Core

3072 GB RAM

103 TB Space

−−

Bandung

336 Core

560 GB RAM

67 TB Space

(15)
(16)

Apa itu Komputasi paralel?

Komputasi Serial:

Komputer desktop

konvensional memiliki Central Processing Unit tunggal (CPU) dan komputasi dilakukan

dan komputasi dilakukan

dengan memecah problem menjadi serangkaian perintah diskrit.

Perintah di eksekusi oleh

komputer satu persatu, karena hanya satu perintah yang

(17)

Apa itu Komputasi paralel?

Komputasi Paralel:

Sedangkan Hardware High

Performance Computing terdiri dari beberapa CPU dan dikonfigurasi untuk menjalankan perhitungan paralel.

menjalankan perhitungan paralel.

Setiap problem harus dipecah menjadi

bagian-bagian diskrit yang dapat dikomputasi secara konkuren

Setiap bagian kemudian dipecah

menjadi serangkaian perintah.

Perintah tersebut dikomputasi secara

(18)

Apa itu Komputasi paralel?

Masalah komputasi yang akan dijalankan di HPC harus

dapat:

Dipisah-pisahkan menjadi potongan-potongan diskrit

pekerjaan yang dapat diselesaikan secara bersamaan;

Mengeksekusi beberapa instruksi program pada setiap saat

Mengeksekusi beberapa instruksi program pada setiap saat

dalam waktu;

Diselesaikan dalam waktu kurang dengan beberapa sumber

komputasi daripada dengan sumber daya komputasi tunggal.

sumber daya komputasi biasanya:

komputer dengan prosesor / core banyak

(19)

Apa itu Komputasi paralel?

Jika anda memiliki aplikasi komputasi favorit

 Satu prosesor akan memberi hasil dalam N jam.  Mengapa tidak menggunakan N prosesor

-- dan mendapat hasil hanya dalam 1 jam?

Konsepnya :

Parallelism = menggunakan beberapa prosesor pada sebuah problem

Dua komponen parallel programming

Komputasi

(20)

A Computer Cluster

Regular PC A computer cluster

Front-end node

Compute-0-0 Compute-0-1 Compute-0-2

1 CPU

(21)

Parallel computing is computing by committee

komputasi paralel: penggunaan beberapa komputer atau

prosesor yang bekerja bersama-sama dalam tugas bersama.

Setiap prosesor bekerja pada bagiannya masing-masing dari

problem

Prosesor diperbolehkan untuk bertukar informasi dengan prosesor Prosesor diperbolehkan untuk bertukar informasi dengan prosesor

lainnya

CPU #1 works on this area of the problem

CPU #3 works on this area

of the problem CPU #4 works on this areaof the problem CPU #2 works on this area

of the problem Grid of Problem to be solved

(22)

Mengapa menggunakan HPC?

Calculation will increasingly

replace experimentation in design

Data + Simulation = Innovation

replace experimentation in design

of useful materials, catalysts, and

drugs, leading to much greater

efficiency and new opportunities

for creativity

(23)
(24)

Mengapa

(25)

Mengapa menggunakan HPC?

 Dunia nyata parallel secara masiv:

 Di dunia nyata, banyak peristiwa yang kompleks dan saling terkait yang terjadi

pada saat yang sama, namun dalam urutan temporal.

 Dibandingkan dengan komputasi serial, komputasi paralel jauh lebih cocok untuk

pemodelan, simulasi dan pemahaman fenomena dunia nyata yang kompleks.

 Misalnya, bayangkan melakukan pemodelan hal-2 berikut secara  Misalnya, bayangkan melakukan pemodelan hal-2 berikut secara

(26)

Mengapa menggunakan HPC?

Misalnya, bayangkan melakukan pemodelan hal-2 berikut

(27)

Menghemat waktu dan/atau uang

Secara teori, menggunakan lebih banyak sumber daya pada

sebuah pekerjaan akan mempersingkat waktu penyelesaian,

dengan potensi penghematan biaya.

(28)

Menghemat

waktu

dan/atau

uang

(29)

Menghemat

waktu

dan/atau

uang

(30)

MEMECAHKAN MASALAH YANG LEBIH BESAR / KOMPLEKS:

 Banyak masalah yang begitu besar dan / atau kompleks yang secara

teknis tidak atau tidak mungkin untuk dipecahkan dengan satu komputer, terutama mengingat memori komputer yang terbatas.

 Contoh:

 Mesin pencari / database pengolahan jutaan transaksi setiap detik  “Masalah yang menjadi tantangan besar”

 “Masalah yang menjadi tantangan besar”

(31)

Grand challenge Problem

Solving grand challenge applications using computer

modeling

,

simulation

and

analysis

Life Sciences Life Sciences

CAD/CAM CAD/CAM

Aerospace Aerospace

Military Applications Digital Biology

(32)
(33)
(34)

Life

(35)
(36)
(37)
(38)

Signal Processing/Quantum Mechanics

Signal Processing/Quantum Mechanics

Convolution model (stencil)

Matrix computations (eigenvalues…) Conjugate gradient methods

Normally not very demanding on latency and bandwidth Some algorithms are embarrassingly parallel

(39)
(40)

Pekerjaan Dilakukan Secara Konkuren

 Sebuah sumber daya komputasi tunggal hanya dapat melakukan satu hal pada

suatu waktu.

 Beberapa sumber daya komputasi dapat melakukan banyak hal secara

bersamaan.

 Contoh: Jaringan Kolaborasi menyediakan tempat global di mana orang-orang dari

(41)

Siapa yang Menggunakan Parallel Computing?

 Science dan Engineering :

 Secara historis, komputasi paralel telah dianggap “komputasi high end”, dan telah

digunakan untuk memodelkan masalah sulit di banyak bidang ilmu pengetahuan dan teknik:

 Atmosphere, Earth, Environment

 Physics - applied, nuclear, particle, condensed matter, high pressure, fusion, photonics  Bioscience, Biotechnology, Genetics

 Chemistry, Molecular Sciences  Geology, Seismology

 Mechanical Engineering - from prosthetics to spacecraft

 Electrical Engineering, Circuit Design, Microelectronics

 Computer Science, Mathematics

(42)

HPC Applications and Major Industries

 Finite Element Modeling

 Auto/Aero

 Fluid Dynamics

 Auto/Aero, Consumer Packaged Goods

Mfgs, Process Mfg, Disaster Preparedness (tsunami)

Imaging

42

 Imaging

Seismic & Medical

 Finance

 Banks, Brokerage Houses (Regression

Analysis, Risk, Options Pricing, What if, …)

 Molecular Modeling

 Biotech and Pharmaceuticals

5 January 2017

Complex Problems, Large Datasets, Long Runs

Complex Problems, Large Datasets, Long Runs

(43)

Divide and Conquer

Says 1 CPU

1,000,000 elements

Numerical processing for 1

element = .1 secs

43

element = .1 secs

One computer will take

100,000 secs = 27.7 hrs

Says 100 CPUs

.27 hr ~ 16 mins

(44)

Life Science Problem – an example of Protein

Folding

 Take a computing year (in serial mode) to do molecular

dynamics simulation for a protein folding problem

5 January 2017

•Excerpted from IBM David Klepacki’s The future of HPC

(45)

Disaster Preparedness

Project LEAD

Severe Weather prediction

(Tornado) – OU leads.

HPC & Dynamically

adaptation to weather

adaptation to weather

forecast

Professor Seidel’s LSU CCT

Hurricane Route Prediction

Emergency Preparedness

Show Movie – HPC-enabled

Simulation

(46)

Cancer Gene-mining

 Unsuccessful on a uni-processor

 Approach

Novel parallel gene-mining algorithms Input from microarray

Retain accuracy

Significantly speed up (superlinear) Significantly speed up (superlinear)

 IBM P5 supercomputer (128 node PPC).

0 20 40 60 80 100Bladder Breast Leukemia Lung Colorectal Lymphoma Melanoma Ovary Pancreas Prostate Renal Mesothelioma

OvaMarker based Selection GeneSetMine based Selection

5 January 2017 46

Time to run the algorithm, keeping number of nodes fixed

0 200 400 600 800 1000 1200

13 39 65 91

Number of processors

(47)

Did you know that Playstation 3 is a

HPC/Supercomputer?

 9 cores/CPUs in one chip.

 Future gaming software is no longer graphic or multimedia only

(48)

Global Climate Modeling Problem

Problem is to compute:

f(latitude, longitude, elevation, tim

e)

temperature, pressure, humidity,

48

temperature, pressure, humidity,

wind velocity

Approach:

Discretize

the domain, e.g., a

measurement point every 10 km

Devise an algorithm to predict

weather at time t+1 given t

C Cox

• Uses:

(49)

Global Climate Modeling Computation

Computational requirements:

To match real-time, need 5x 10

11

flops in 60 seconds = 8

Gflop/s

Weather prediction (7 days in 24 hours)

56 Gflop/s

49

Weather prediction (7 days in 24 hours)

56 Gflop/s

Climate prediction (50 years in 30 days)

4.8 Tflop/s

To use in policy negotiations (50 years in 12 hours)

288

Tflop/s

 To double the grid resolution, computation is at least 8x

 State of the art models require integration of atmosphere, ocean, sea-ice, land models, plus

possibly carbon cycle, geochemistry and more

(50)

Heart Simulation

Problem is to compute blood flow in the heart

Approach:

Modeled as an elastic structure in an incompressible fluid.

The “immersed boundary method” due to Peskin and McQueen.

20 years of development in model

50

20 years of development in model

Many applications other than the heart: blood clotting, inner ear, paper making, embryo growth, and others

Use a regularly spaced mesh (set of points) for evaluating the fluid flow

Uses

Current model can be used to design artificial heart valves Can help in understand effects of disease (leaky valves)

Related projects look at the behavior of the heart during a heart attack

Ultimately: real-time clinical work

(51)

Parallel computing: Web searching

51

• Functional parallelism: crawling, indexing, sorting

• Parallelism between queries: multiple users

C Cox

• Finding information amidst junk

• Preprocessing of the web data set to help find information

(52)

Parallel Programming: Decomposition Techniques

Functional Decomposition (Functional Parallelism)

 Decomposing the problem into different tasks which can be distributed

to multiple processors for simultaneous execution

 Good to use when there is not static structure or fixed determination of

number of calculations to be performed

52

Domain Decomposition (Data Parallelism)

 Partitioning the problem's data domain and distributing portions to

multiple processors for simultaneous execution

 Good to use for problems where:

 data is static (e.g. solving large matrix or finite difference or finite element calculations)  dynamic data structure tied to single entity where entity can be separated

 domain is fixed but computation within various regions of the domain is dynamic (fluid vortices models)

(53)

Siapa yang Menggunakan Parallel Computing?

Bioscience, Biotechnol ogy, Genetics

Atmosphere, Earth, En vironment

(54)
(55)

Siapa yang Menggunakan Parallel Computing?

Industrial and Commercial

 Aplikasi-aplikasi berikut memerlukan pengolahan data dalam jumlah besar dengan cara yang canggih.

Big Data, databases, data

mining Financial and economic modeling mining Financial and economic modeling Oil exploration Management of national and multi-national corporations

Web search engines, web based business services

Advanced graphics and virtual reality, particularly in the

entertainment industry Medical imaging and

(56)

Top Ten Most Powerful Computers http://www.top500.org)

# Site System Cores (TFlop/s)Rmax (TFlop/s)Rpeak Power (kW)

1 National Supercomputing Center

in WuxiChina Sunway TaihuLight1.45GHz, Sunway NRCPC- Sunway MPP, Sunway SW26010 260C 10,649,600 93,014.6 125,435.9 15,371 2 National Super Computer Center

in GuangzhouChina Tianhe-2 (MilkyWay-2)12C 2.200GHz, TH Express-2, Intel Xeon Phi 31S1P- TH-IVB-FEP Cluster, Intel Xeon E5-2692 NUDT 3,120,000 33,862.7 54,902.4 17,808 3 DOE/SC/Oak Ridge National

LaboratoryUS Titaninterconnect, NVIDIA K20x- Cray XK7 , Opteron 6274 16C 2.200GHz, Cray Gemini Cray Inc. 560,640 17,590.0 27,112.5 8,209

LaboratoryUS interconnect, NVIDIA K20x Cray Inc.

4 DOE/NNSA/LLNLUS Sequoia- BlueGene/Q, Power BQC 16C 1.60 GHz, CustomIBM 1,572,864 17,173.2 20,132.7 7,890 5 DOE/SC/LBNL/NERSC US Cori- Cray XC40, Intel Xeon Phi 7250 68C 1.4GHz, Aries

interconnect Cray Inc. 622,336 14,014.7 27,880.7 3,939 6 Joint Center for Advanced HPC

Japan Oakforest-PACS68C 1.4GHz, Intel Omni-Path- PRIMERGY CX1640 M1, Intel Xeon Phi 7250 Fujitsu 556,104 13,554.6 24,913.5 2,719 7 RIKEN (AICS)Japan K computer, SPARC64 VIIIfx 2.0GHz, Tofu interconnect Fujitsu 705,024 10,510.0 11,280.4 12,660 8 Swiss National Supercomputing

Centre (CSCS)Switzerland Piz Daintinterconnect , NVIDIA Tesla P100- Cray XC50, Xeon E5-2690v3 12C 2.6GHz, Aries Cray Inc. 206,720 9,779.0 15,988.0 1,312 9 DOE/SC/Argonne National

LaboratoryUS Mira- BlueGene/Q, Power BQC 16C 1.60GHz, Custom IBM 786,432 8,586.6 10,066.3 3,945 10 DOE/NNSA/LANL/SNLUS Trinity- Cray XC40, Xeon E5-2698v3 16C 2.3GHz, Aries

(57)
(58)
(59)

1984 Computer Food Chain

Mainframe

Vector Supercomputer

(60)

1994 Computer Food Chain

Mini Computer (hitting wall soon)

Mainframe

Vector Supercomputer MPP

(61)
(62)

CLUSTERING OF COMPUTERS

FOR COLLECTIVE COMPUTING: TRENDS

?

(63)

21 00

2 1 00 2 1 00 2 1 00 2 1 00 2 1 00 2 1 00 2 1 00 2 1 00

P E R F O

Computing Platforms Evolution

Computing Platforms Evolution

Breaking Administrative Barriers

Breaking Administrative Barriers

Administrative Barriers

?

Desktop

(Single Processor?) SuperComSMPs orputers

Local

Cluster Cluster/GridGlobal

O R M A N C E Inter Planet Cluster/Grid ?? Individual Group Department Campus Sta te National Globe Inte r Plane t Universe

Administrative Barriers

(64)

Cluster Computer and its

Components

(65)

Clustering gained momentum when 3 technologies

converged:

1. Very HP Microprocessors

 workstation performance = yesterday supercomputers

2. High speed communication

2. High speed communication

 Comm. between cluster nodes >= between processors in an SMP.

(66)

Parallel architectures (1)

 Vector machines

 CPU processes multiple data sets  shared memory

 advantages: performance, programming difficulties  issues: scalability, price

 examples: Cray SV, NEC SX, Athlon3/d, Pentium- IV/SSE/SSE2

 Massively parallel processors (MPP)

 large number of CPUs  distributed memory

 advantages: scalability, price

 issues: performance, programming difficulties

(67)

Parallel architectures (2)

 Symmetric Multiple Processing (SMP)

 two or more processors  shared memory

 advantages: price, performance, programming difficulties  issues: scalability

examples: UltraSparcII, Alpha ES, Generic Itanium, Opteron, Xeon, …

 examples: UltraSparcII, Alpha ES, Generic Itanium, Opteron, Xeon, …

 Non Uniform Memory Access (NUMA)

 Solving SMP’sscalability issue  hybrid memory model

 advantages: scalability

(68)

Clusters

Cluster consists of:

 Nodes

 Network

 OS

 Cluster middleware

 Standard components  Standard components

(69)

Cluster Architecture

Sequential Applications

Parallel Applications

Parallel Programming Environment

Cluster Middleware

(Single System Image and Availability Infrastructure) Sequential Applications

Sequential Applications

Parallel Applications

Parallel Applications

(Single System Image and Availability Infrastructure)

Cluster Interconnection Network/Switch

(70)

Cluster Components...1a

Nodes

Multiple High Performance Components:

PCs

Workstations

Workstations

SMPs (CLUMPS)

Distributed HPC Systems leading to

Metacomputing

They can be based on different

(71)

Cluster Components...1b

Processors

 There are many (CISC/RISC/VLIW/Vector..)

 Intel: Pentiums, Xeon, Merceed….  Sun: SPARC, ULTRASPARC

 HP PA

IBM RS6000/PowerPCIBM RS6000/PowerPC  SGI MPIS

 Digital Alphas

 Integrate Memory, processing and networking into a single

chip

IRAM (CPU & Mem):

(http://iram.cs.berkeley.edu)

(72)

Cluster Components…2

OS

State of the art OS:

 Linux (Beowulf)

 Microsoft NT (Illinois HPVM)  SUN Solaris (Berkeley NOW)  SUN Solaris (Berkeley NOW)  IBM AIX (IBM SP2)

 HP UX (Illinois - PANDA)

 Mach (Microkernel based OS) (CMU)

 Cluster Operating Systems (Solaris MC, SCO Unixware, MOSIX

(academic project)

(73)

Cluster Components…3

High Performance Networks

Ethernet (10Mbps),

Fast Ethernet (100Mbps),

Gigabit Ethernet (1Gbps)

SCI (Dolphin - MPI- 12micro-sec latency)

ATM

Myrinet (1.2Gbps)

(74)

Cluster Components…4

Network Interfaces

Network Interface Card

Myrinet has NIC

User-level access support

User-level access support

Alpha 21364 processor integrates

(75)

Cluster Components…

5 Communication Software

 Traditional OS supported facilities (heavy weight due

to protocol processing)..

Sockets (TCP/IP), Pipes, etc.  Light weight protocols (User Level)  Light weight protocols (User Level)

Active Messages (Berkeley) Fast Messages (Illinois)

U-net (Cornell) XTP (Virginia)

 System systems can be built on top of the above

(76)

Cluster Components…6a

Cluster Middleware

Resides Between OS and Applications and

offers in infrastructure for supporting:

Single System Image (SSI)

System Availability (SA)

SSI makes collection appear as single

machine (globalised view of system

resources). Telnet cluster.myinstitute.edu

(77)

Cluster Components…6b

Middleware Components

Hardware

 DEC Memory Channel, DSM (Alewife, DASH) SMP Techniques

OS / Gluing Layers

OS / Gluing Layers

 Solaris MC, Unixware, Glunix)

Applications and Subsystems

 System management and electronic forms

 Runtime systems (software DSM, PFS etc.)

 Resource management and scheduling (RMS):

(78)

Cluster Components…7a

Programming environments

Threads (PCs, SMPs, NOW..)

POSIX Threads Java Threads Java Threads

MPI

Linux, NT, on many Supercomputers

PVM

(79)

Cluster Components…7b

Development Tools ?

Compilers

 C/C++/Java/ ;

 Parallel programming with C++ (MIT Press book)

RAD (rapid application development tools)..

RAD (rapid application development tools)..

GUI based tools for PP modeling

Debuggers

(80)

Cluster Components…8

Applications

Sequential

Parallel / Distributed (Cluster-aware app.)

Grand Challenging applications

Grand Challenging applications

Weather Forecasting

Quantum Chemistry

Molecular Biology Modeling

Engineering Analysis (CAD/CAM)

……….

(81)

Classification

(82)

Clusters Classification..1

Based on Focus (in Market)

High Performance (HP) Clusters

Grand Challenging Applications

High Availability (HA) Clusters

(83)

Clusters Classification..2

Based on Workstation/PC Ownership

Dedicated Clusters

Non-dedicated clusters

Adaptive parallel computing

Also called Communal

(84)

Clusters Classification..3

Based on Node Architecture..

Clusters of PCs (CoPs)

Clusters of Workstations (COWs)

Clusters of SMPs (Symmetric

(85)

Clusters Classification..4

Based on Node OS Type..

Linux Clusters (Beowulf)

Solaris Clusters (Berkeley NOW)

NT Clusters (HPVM)

NT Clusters (HPVM)

AIX Clusters (IBM SP2)

SCO/Compaq Clusters (Unixware)

…….Digital VMS Clusters, HP

(86)

Clusters Classification..5

Based on node components architecture &

configuration (Processor Arch, Node Type:

PC/Workstation.. & OS: Linux/NT..):

Homogeneous Clusters

Homogeneous Clusters

 All nodes will have similar configuration

Heterogeneous Clusters

Nodes based on different processors

(87)

Clusters Classification..6a

Dimensions of Scalability & Levels of Clustering

Network

Technology

(1)

(3)

Campus Enterprise

Public Metacomputing (GRID)

(88)

Clusters Classification..6b

Levels of Clustering

Group Clusters (#nodes: 2-99)

 (a set of dedicated/non-dedicated computers - mainly connected by SAN like

Myrinet)

 Departmental Clusters (#nodes: 99-999)  Organizational Clusters (#nodes: many 100s)  (using ATMs Net)

 Internet-wide Clusters=Global Clusters: (#nodes: 1000s to many millions)

 Metacomputing

 Web-based Computing  Agent Based Computing

(89)

Major issues in cluster design

Size Scalability (physical &

application) Enhanced Availability (failure management)

Single System Image (look-and- Fast Communication (networks &

Single System Image

(look-and-feel of one system) Fast Communication (networks & protocols)

Load Balancing (CPU, Net,

Memory, Disk) Security and Encryption (clusters of clusters)

Distributed Environment (Social

issues) Manageability (admin. And control)

Programmability (simple API if

(90)

What Next ??

Clusters of Clusters (HyperClusters) Global Grid

(91)
(92)
(93)

What is Grid ?

 An infrastructure that couples

Computers (PCs, workstations, clusters, traditional

supercomputers, and even laptops, notebooks, mobile computers, PDA, and so on)

Software ? (e.g., renting expensive special purpose

applications on demand) applications on demand)

Databases (e.g., transparent access to human genome

database)

Special Instruments (e.g., radio telescope--SETI@Home

Searching for Life in galaxy, Austrophysics@Swinburne for pulsars)

People (may be even animals who knows ?)

 across the local/wide-area networks (enterprise, organisations, or Internet)

(94)

Conceptual view of the Grid

Leading to Portal (Super)Computing Leading to Portal (Super)Computing

(95)

Grid Application-Drivers

Old and New applications getting enabled due to

coupling of computers, databases, instruments, people,

etc:

(distributed) Supercomputing Collaborative engineering high-throughput computing

large scale simulation & parameter studies

Remote software access / Renting Software Data-intensive computing

(96)

Grid Components

Development Environments and Tools

Languages Libraries Debuggers Monitoring Resource Brokers … Web tools

Applications and Portals

Prob. Solving Env.

Scientific Engineering Collaboration … Web enabled Apps Grid Apps.

Grid Tools

Grid Fabric

Networked Resources across Organisations

Computers Clusters Storage Systems Data Sources Scientific Instruments

Local Resource Managers

Operating Systems Queuing Systems TCP/IP & UDP

Libraries & App Kernels …

Distributed Resources Coupling Services

(97)

Many GRID Projects and Initiatives

 PUBLIC FORUMS

 Computing Portals  Grid Forum

 European Grid Forum  IEEE TFCC!

 GRID’2000 and more.

Europe UNICORE MOL METODIS Globe Poznan Metacomputing CERN Data Grid MetaMPI USA Globus Legion JAVELIN AppLes NASA IPG Condor

Harness  GRID’2000 and more.

 Public Grid Initiatives

 Distributed.net  SETI@Home

 Compute Power Grid

 Japan

 Ninf

 Bricks

 and many more...

MetaMPI DAS

JaWS

and many more...

Australia

Nimrod/G

EcoGrid and GRACE DISCWorld Condor Harness NetSolve NCSA Workbench WebFlow EveryWhere

(98)

Literature on Cluster

Literature on Cluster

(99)

Reading Resources..1

Internet & WWW

Computer Architecture:

http://www.cs.wisc.edu/~arch/www/

Linux Parallel Procesing

 http://yara.ecn.purdue.edu/~pplinux/Sites/

Solaris-MC

Solaris-MC

http://www.sunlabs.com/research/solaris-mc

Microprocessors: Recent Advances

 http://www.microprocessor.sscc.ru

Beowulf:

 http://www.beowulf.org

Metacomputing

(100)

Reading Resources..2

Books

In Search of Cluster

 by G.Pfister, Prentice Hall (2ed), 98

High Performance Cluster Computing

Volume1: Architectures and Systems

Volume1: Architectures and Systems

Volume2: Programming and Applications

 Edited by Rajkumar Buyya, Prentice Hall, NJ, USA.

Scalable Parallel Computing

(101)

Cluster Computing Forum

IEEE Task Force on Cluster Computing

(TFCC)

(102)

TFCC Activities...

 Network Technologies  OS Technologies

Parallel I/O

Programming EnvironmentsProgramming EnvironmentsJava Technologies

 Algorithms and Applications  >Analysis and Profiling

 Storage Technologies

(103)

TFCC Activities...

 High Availability  Single System ImagePerformance EvaluationSoftware EngineeringSoftware EngineeringEducation

 Newsletter  Industrial Wing

 TFCC Regional Activities

(104)

TFCC Activities...

 Mailing list, Workshops, Conferences, Tutorials, Web-resources etc.

Resources for introducing subject in senior undergraduate and

graduate levels.

 Tutorials/Workshops at IEEE Chapters..  ….. and so on.

FREE MEMBERSHIP, please join!Visit TFCC Page for more details:

(105)

Referensi

Dokumen terkait

Tujuan Penelitian : Untuk mengetahui factor-faktor yang berhubungan dengan kepatuhan pasien diabetes mellitus dan hipertensi dalam mengikuti kegiatan prolanis di

Pada Wikipedia dengan alamat http://id.wikipedia.org komunikasi antar budaya adalah komunikasi yang terjadi di antara orang-orang yang memiliki

Nilai perusahaan merupakan penilaian investor terhadap tingkat keberhasilan perusahaan dalam mengelola sumber daya pada akhir tahun berjalan yang tercermin pada harga

Evasari Hutajulu: Tata Cara Penyelesaian Kredit Macet Pada PT... Evasari Hutajulu: Tata Cara Penyelesaian Kredit Macet

Dinamika sosial politik Indonesia yang juga berdampak pada perubahan kurikulum merupakan suatu bentuk penyempurnaan dalam bidang pendidikan untuk meningkatan

Pemerintah Indonesia dan masyarakat Indonesia dalam menyikapi persoalan LGBT ini, dengan dalih pembelaan HAM, ada yang mendukung LGBT dan berupaya agar diterima

Selain oleh PPID Perpustakaan Nasional RI, pelatihan perpustakaan juga banyak diselenggarakan oleh perguruan tinggi dan lembaga pemerintah lainnya, seperti: (1) Program

Sehubungan dengan Berita Acara Evaluasi Dokumen Penawaran Nomor : BA.01/BOR.253.LPSE/ULP_POKJA III/LMD/V/2015 tanggal 15 Mei 2015 untuk Pekerjaan Lanjutan