C2115 Practical Introduction to Supercomputing 2nd Lesson -1C2115
Practical Introduction to
Supercomputing
Petr Kulhánek, Jakub Štěpán
kulhanek@chemi.muni.cz
National Centre for Biomolecular Research, Faculty of Science
Masaryk University, Kotlářská 2, CZ-61137 Brno
CZ.1.07/2.2.00/15.0233
2. lekce
C2115 Practical Introduction to Supercomputing 2nd Lesson -2-
Contents
 History, usage and future of computing
 Real problems examples
Laboratory of Computational Chemistry projects
 CZ Supercomputer centres overview
MetaCentrum, CERIT-SC, IT4 Innovation
 Foreigin supercomputer centres
centres accessible from applications from CZ, Top500
 Exercise
C2115 Practical Introduction to Supercomputing 2nd Lesson -3-
History
http://en.wikipedia.org/wiki/History_of_computing_hardware
1985 Cray 2 1,9 GFLOPS
http://www.root.cz
Pavel Tišnovský, Unixové vykopávky1800 punched cards beginings
1946 ENIAC
1947 transistor discovery
1971 Intel 4004 (4 bit)
1974 Intel 8080 (8 bit)
1976 Intel 8086 (16 bit)
1985 Intel 80386 (32 bit)
2001 IA-64 (64 bit)
2003 AMD64/EM64T (64 bit)
proprietary vector CPU
2010 Intel Core i7 980X: @3,33 GHz
(6C/12T, Turbo@3,46 GHz): 109 GFLOPS
source: wikipedia.org, intel.com
C2115 Practical Introduction to Supercomputing 2nd Lesson -4Computing
devices usage
Computing devices (computers) influenced all human activities and became undispensable
part of our life. Rapid development in last 20 years caused that. Computing is used in
entertainment, processing and concumtion of information.
Computing devices (at most supercomputers) are used to solve demanding numerical
problems as:
 Wheather forecast simulations, climete and geologycal simulations (floods, tsunami,
earthquake)
 New material and drug design
 Economy modelling
 Scientific-technical calculations (chemistry, fyzics, mathematics)
 Military purposes (nuclear veapons simulations)
 .....
C2115 Practical Introduction to Supercomputing 2nd Lesson -5-
Future
?
http://www.humanbrainproject.eu/
Human brain simulator:
... quantum computers, ...
Maasive usage GPGPU ...
C2115 Practical Introduction to Supercomputing 2nd Lesson -6Laboratory
of
Computational Chemistry
 Project overview
C2115 Practical Introduction to Supercomputing 2nd Lesson -7Group
of Computational Chemistry
prof. RNDr. Jaroslav Koča, DrSc.
1 professor
5 senior researchers
2 post-doc students
11 doktoral students
7 bachelor and magister students
Mgr. Zdeněk Kříž, Ph.D.
E-mail: zdenek.kriz@ceitec.muni.cz
Expertise: QM, MD, Docking
Mgr. Martin Prokop, Ph.D.
E-mail: martin.prokop@ceitec.muni.cz
Expertise: Software dev, Dockig
RNDr. Petr Kulhánek, Ph.D.
E-mail: petr.kulhanek@ceitec.muni.cz
Expertise: QM, QM/MM, MD, Free Energy
RNDr. Radka Svobodová, Ph.D.
E-mail: radka.svobodova@ceitec.muni.cz
Expertise: Chemo and Bioinformatics
RNDr. Robert Vácha, Ph.D.
E-mail: robert.vacha@ceitec.muni.cz
Expertise: MD, MC, Coarse Grain, Free Energy
Mgr. Stanislav Kozmon, Ph.D.
E-mail: stano@chemi.muni.cz
Expertise: QM, QM/MM
C2115 Practical Introduction to Supercomputing 2nd Lesson -8Computational
chemistry
Computational Chemistry
is a branch of chemistry that uses computer simulation to assist in
solving chemical problems. It uses methods of theoretical
chemistry, incorporated into efficient computer programs, to
calculate the structures and properties of molecules and solids.
While computational results normally complement the information
obtained by chemical experiments, it can in some cases predict
hitherto unobserved chemical phenomena. It is widely used in the
design of new drugs and materials.
www.wikipedia.org
C2115 Practical Introduction to Supercomputing 2nd Lesson -9Interdisciplinary
field
C2115 Practical Introduction to Supercomputing 2nd Lesson -10What
we do and offer …
• expertize in field of computational chemistry
Molecular dynamics simulations, docking, chemo a bioinformatics, bonding
energy, quantum mechanics, coarse grained models …
• Development of computational tools, approaches and software
in silico drug design and protein mutagenesis, free energy calulations, fast
charge calculations, virtual screening, tunnel and cavity prediction in
biomolecules …
• studium systémů různých velikostí
lectins enzymes large biomolecular complexesSmall complexes
C2115 Practical Introduction to Supercomputing 2nd Lesson -11Theory
levels
Quantum mechanics Molecular mechanics Coarse-grained mechanics
atomic resolution bead resolution
reactivity Domains movement, folding
atomic resolution bead resolutionatomic resolution bead resolution
Conformational changes
Up to 1'000 atoms * Up to 1'000'000 beads *Up to 1'000'000 atoms *
Up to 100 ps * Up to ms *Up to 1 ms *
C2115 Practical Introduction to Supercomputing 2nd Lesson -12-
Projects
(bio)molecular systems studies
C2115 Practical Introduction to Supercomputing 2nd Lesson -13Quantum
chemical calculations
)()(ˆ rr kkk
EH  
Time independent Schrödinger equation
Metods
Formal scaling HF CI metods MP metods CC metods
N4 -> N2 -> N1 HF,DFT
N5 MP2 CC2 (iterative)
N6 CISD MP3, MP4(SDQ) CCSD (iterative)
N7 MP4 CCSD(T), CC3 (iterative)
N8 CISDT MP5 CCSDT
N9 MP6
N10 CISDTQ MP7 CCSDTQ (iterative)
Scaling, time demands: http://en.wikipedia.org/wiki/Time_complexity
HF - Hartree–Fock metod, DFT – density functional theory,
CI – configuration interaction methods, MP - Møller–Plesset perturbation teory,
CC – coupled cluster metod, N – base function number
Jensen, F. Introduction to computational chemistry; 2nd ed.; John Wiley & Sons:
Chichester, England; Hoboken, NJ, 2007.
C2115 Practical Introduction to Supercomputing 2nd Lesson -14Quantum
chemical calculations
supramoleculr complexes
Sacharides binding abilities
C2115 Practical Introduction to Supercomputing 2nd Lesson -15Molecular
mechanics
Schrödinger equation => quantum mechanics view
Bonding term Non-bonding termsClassical physics => mechanical view
Approximation uses classical physics not considering
explicit electron movement (electrons are included in
empirical parameters)
Formal scaling: N2 -> N log2N
N – atom number
C2115 Practical Introduction to Supercomputing 2nd Lesson -16Molecular
dynamics
iii
m= aF
Equation system of second order
needs numerical solution
Discrete molecular movement in short time steps
1 fs
Determined by fastest movement (bond vibration)
Integration imperfections are corrected by usage of thermostats and barostats, that ensure
given simulation conditions.
F
R
R




)(E
II. Newton law of motion (law of force)
2
2
dt
d i
i
r
a 
Typical integration step
C2115 Practical Introduction to Supercomputing 2nd Lesson -17Conformational
changes, transport
endonuclease BsoBI opening
• Conformational changes and transport are important for biomolecular system
function
Iontransport in DNA qadruplexes
C2115 Practical Introduction to Supercomputing 2nd Lesson -18Enzyme
reaction mechanisms QM/MM
QM
MM
C2115 Practical Introduction to Supercomputing 2nd Lesson -19Virtual
screening
Motivation
Software tools
Cíl
Performance
Lectin PA-IIL
• Autodock Vina
• Heterogennous computational sources
• One docking cca 1-10 min/ligand
• ca 900 parallel tasks
• Search speed cca 250 000 ligands per day
C2115 Practical Introduction to Supercomputing 2nd Lesson -20-
Contacts
Laboratory of computational chemistry
National Centre for Biomolecular Research, UKB, Pavilon A4
http://lcc.ncbr. muni.cz
Seminars LCC group are on Thursday at 10 A.M. in room 2.11/A4.
C2115 Practical Introduction to Supercomputing 2nd Lesson -21Computing
centres in CZ
 MetaCentrum
 CERIT-SC
 MetaCentrum VO
 IT4 Innovation
C2115 Practical Introduction to Supercomputing 2nd Lesson -22-
MetaCentrum
http://www.metacentrum.cz/
Long-term goal of the MetaCentrum project is operation and coordination of distributed
computing and data storage infrastructure accompanied by an appropriate support
environment and continual expansion of available computational capacities.
The main aim of the project is constitution of a virtual computer that allows effective
utilization of installed facilities in the frame of supercomputing project and solving tasks
whose memory and/or CPU requirements exceeds possibility of individual single
supercomputing centers.
MetaCentrum is CESNET activity.
C2115 Practical Introduction to Supercomputing 2nd Lesson -23-
CERIT-SC
Center CERIT-SC (CERIT Scientific Cloud) is national center providing flexible storage and
computional facilities and conected services, including support of their experimental
support. Moreover center does research in field of flexible e-infrastructure and cooperates
on research with users.
Center CERIT-SC developed from Supercomputing center Brno (SCB), that is part of
Institute Ústav výpočetní techniky (ÚVT) of Masaryk University (MU).
Center CERIT-SC plans to provide more than 3500 cores and 3,5 PB storage by 2013. These
resources will be continuously put into operation and are accessible through National Grid
Infrastructure, cloud and other interfaces.
http://www.cerit-sc.cz/
C2115 Practical Introduction to Supercomputing 2nd Lesson -24MetaCentrum
VO
• National Grid Infrastructure
• OS Debian
• ca 2500 CPU cores
• CEITEC/NCBR own resources cca 850 CPU cores
• 3 x 100 TB storage capacity
• cca 3 TB per user
Associates resources provided by MetaCentre, project CERIT-SC and other partners.
http://metavo.metacentrum.cz/
 Account may be provided to any CZ university
student.
 Account is not limited to particular project and is
granted for one year.
 Extension is done based on year report.
C2115 Practical Introduction to Supercomputing 2nd Lesson -25MetaCentrum
VOMetaCentrum VO
• National Grid Infrastructure
• OS Debian
• ca 2500 CPU cores
• CEITEC/NCBR own resources cca 850 CPU cores
• 3 x 100 TB storage capacity
• cca 3 TB per user
Associates resources provided by MetaCentre, project CERIT-SC and other partners.
http://metavo.metacentrum.cz/
 Account may be provided to any CZ university
student.
 Account is not limited to particular project and is
granted for one year.
 Extension is done based on year report.
C2115 Practical Introduction to Supercomputing 2nd Lesson -26IT4
Innovation
IT4Innovations project goals are to build national centre of excelent research in information
technologies. Part of the project is purchase of supercomputer, that is to be started by year
2014.
Centre basis will be in computing, that is devided into three interconnected research fields:
 IT4People (Information Technology for People) – research directed to quality of life
improvement using modern information technologies.
 SC4Industry (Supercomputing for Industry) – supercomputing in solving industry
problems, modelling in field of nature science and nanotechnologies (shape
optimisations, material design, biomechanical simulations, ...).
 Theory4IT (Theory for Information Technology) – field directed to basic research,
development of new calculation methods (data mining, anthill theory).
Project is supported by five subjects: Vysoká škola báňská-Technická univerzita Ostrava, Ostravská univerzita v
Ostravě, Slezská univerzita v Opavě, Vysoké učení technické v Brně a Ústav geoniky AV ČR.
Current call: to 4. 3 2013
http://www.it4i.cz/
C2115 Practical Introduction to Supercomputing 2nd Lesson -27Foreign
computing
centres
 PRACE
 TOP500
C2115 Practical Introduction to Supercomputing 2nd Lesson -28-
PRACE
http://www.prace-ri.eu/
Project Types:
 Multi-year Access is available to major European projects or infrastructures that can
benefit from PRACE resources and for which Project Access is not appropriate.
 Project Access is intended for individual researchers and research groups including
multi-national research groups and has a one year duration. Calls for Proposals for
Project Access are issued twice yearly (February and September).
 Preparatory Access is intended for resource use required to prepare proposals for
Project Access. Applications for Preparatory Access are accepted at any time.
Next call: 13. FEB – 26. MAR 2013
PRACE: Partnership for Advanced Computing in Europe
C2115 Practical Introduction to Supercomputing 2nd Lesson -29PRACE
- členové
Austria: JKU - Johannes Kepler University of Linz
Belgium: DGO6-SPW - Direction générale opérationnelle de l’Économie, de l’Emploi et de la Recherche – Service Public de Wallonie
Bulgaria: NCSA - Executive agency "Electronic communication networks and information systems"
Cyprus: CaSToRC – Computation-based Science and Technology Research Center, The Cyprus Institute
Czech Republic: VŠB - Technical University of Ostrava
Denmark: DeIC - Danish e-Infrastructure Cooperation
Finland: CSC - IT Center for Science Ltd.
France: GENCI - Grand Equipement National de Calcul Intensif
Germany: GCS - GAUSS Centre for Supercomputing e.V
Greece: GRNET - Greek Research and Technology Network S.A.
Hungary: NIIFI - National Information Infrastructure Development Institute
Ireland: ICHEC - Irish Centre for High-End Computing
Israel: IUCC - Inter-University Computation Center
Italy: CINECA - Consorzio Interuniversitario
Norway: SIGMA – UNINETT Sigma AS – The Norwegian Metacenter for Computational Science
The Netherlands: SURFSARA: SARA Computing and Networking Services
Poland: PSNC – Instytut Chemii Bioorganicznej Pan – Institute of Bioorganic Chemistry – Poznan Supercomputing and Networking
Center
Portugal: Universidade de Coimbra
Serbia: IPB - Institute of Physics Belgrade
Slovenia: ULFME - University of Ljubljana, Faculty of Mechanical Engineering
Spain: BSC – Barcelona Supercomputing Center – Centro Nacional de Supercomputación
Sweden: Vetenskapsrådet – Swedish Research Council
Switzerland: ETH – Eidgenössische Technische Hochschule Zürich – Swiss Federal Institute of Technology, Zürich
Turkey: UYBHM – Ulusal Yuksek Basarimli Hesaplama Merkezi, Istanbul Technical University – National Center for High Performance
Computing
UK: EPSRC – The Engineering and Physical Sciences Research Council
C2115 Practical Introduction to Supercomputing 2nd Lesson -30PRACE
Summer of HPC
http://summerofhpc.prace-ri.eu/
Summer of HPC is a PRACE programme that offers summer placements at HPC centres
across Europe. Up to twenty top applicants from across Europe will be selected to
participate. Participants will spend two months working on projects related to PRACE
technical or industrial work to produce a visualisation or video. The programme will run
from July 1st, to August 30th 2013 and will include a kick-off training week.
Flights, accommodation & a stipend will be provided to all successful applicants; all you
need to bring is your interest in computing and some enthusiasm! Prizes will be awarded
for the best participants.
Call: 25. JAN – 17. FEB 2013
C2115 Practical Introduction to Supercomputing 2nd Lesson -31PRACE
Summer of HPC
http://summerofhpc.prace-ri.eu/
Summer of HPC is a PRACE programme that offers summer placements at HPC centres
across Europe. Up to twenty top applicants from across Europe will be selected to
participate. Participants will spend two months working on projects related to PRACE
technical or industrial work to produce a visualisation or video. The programme will run
from July 1st, to August 30th 2013 and will include a kick-off training week.
Flights, accommodation & a stipend will be provided to all successful applicants; all you
need to bring is your interest in computing and some enthusiasm! Prizes will be awarded
for the best participants.
Call: 25. JAN – 17. FEB 2013
C2115 Practical Introduction to Supercomputing 2nd Lesson -32-
TOP500
TOP500 benchmark
Our simple TOP500 approach does not define “supercomputer” as such, but we use a
benchmark to rank systems and to decide on whether or not they qualify for the TOP500
list. The benchmark we decided on was Linpack, which means that systems are ranked only
by their ability to solve a set of linear equations, A x = b, using a dense random matrix A.
http://www.top500.org/
TOP500 is project, that lists 500 most powerfull computers on the Earth.
November 2012
C2115 Practical Introduction to Supercomputing 2nd Lesson -33TOP500
– Topology
symmetric multiprocessing
(CPU have shared memory)
massively parallel processing
(CPU have own memory)
(node has more CPU then is number of nodes)
(node has less CPU cores
then is number of nodes)
C2115 Practical Introduction to Supercomputing 2nd Lesson -34TOP500
– CPU architecture
C2115 Practical Introduction to Supercomputing 2nd Lesson -35TOP500
– Usage types
C2115 Practical Introduction to Supercomputing 2nd Lesson -36TOP500
– Accelerators/Coprocessors
C2115 Practical Introduction to Supercomputing 2nd Lesson -37TOP500
– Accelerators/Coprocessors
C2115 Practical Introduction to Supercomputing 2nd Lesson -38K
– computer, 3rd place
http://www.youtube.com/watch?v=UJPsIu9OaTc
C2115 Practical Introduction to Supercomputing 2nd Lesson -39Exercise
LI.1
1. How many times longer will be energy calculation of benzene molecule by quantum
chemistry method CCSD(T) between base aug-cc-pVDZ and aug-cc-pVTZ? Base function
number is 192 in aug-cc-pVDZ and 414 in aug-cc-VTZ.
2. Kow many days will take simulation of celulose fragment with length 1 ms in explicit
solvent box with total atom number 408609, on graphics card GTX680 in NPT
conditions? Use data provided on:
Http://ambermd.org/gpus/benchmarks.htm#Benchmarks
3. In 2009 average energy consumption on one CZ inhabitant was 1398 kWh (Energetic
regulative office statistics). How many people would take same year energy as
consumes most powerfull computer Titan with performance of cca 17 PFLOPS?