Adventures in Cyberinfrastructure - Clemson University

longtermsingularInternet και Εφαρμογές Web

4 Δεκ 2013 (πριν από 4 χρόνια και 27 μέρες)

68 εμφανίσεις

www.nanoHUB.org

NCN

1

CI Days, Clemson University, May 20, 2008


Mark Lundstrom

Network for Computational Nanotechnology

Discovery Park, Purdue University

West Lafayette, IN

Adventures

in

Cyberinfrastructure:


observations of an accidental tourist

www.nanoHUB.org

NCN

2

some special people

Gerhard Klimeck, Michael McLennan, George Adams,
and Gerry McCartney (Purdue University)

Jim Bottum, Sebastien Goasguen, Krishna Madhavan,
(Clemson University)

José A.B. Fortes (Univ. of Florida)

Nirav Kapadia (Unisys)

+

the Purdue University leadership and NSF program
managers

www.nanoHUB.org

NCN

3

nanoelectronic devices and materials

D

S

G


HfO
2

10 nm SiO
2

p
++

Si

S


D


Al

Gate

SWNT

carbon nanotube
electronics

NW/NT composites

CoFe (2.5)

Ru (0.85)

Insulator

CoFeB (3)

CoFeB (3)

MgO (0.85)

spin torque devices

nanowire PV

nanowire
bio
-
sensors

molecular electronics

www.nanoHUB.org

NCN

4

why I compute

“The purpose of computing is insight
-

not numbers.”






-
Richard Hamming


to develop understanding



to interpret experiments



to explore new devices



to set the stage for more serious simulations

www.nanoHUB.org

NCN

5

educators

experimentalists

designers

students

computational science and engineering

algorithms

HPC

simulation/

CAD

theorists

modelers

CSE

‘closer to the solution’

‘closer to the problem’

www.nanoHUB.org

NCN

6

the nanoHUB story

Nirav Kapadia, Purdue University

1991
-

2001

www.nanoHUB.org

NCN

7

PUNCH (1994
-
2005)


PUNCH v.4


middleware

gridware


Software applications


-
Unix


-
text
-
based / forms
-
based


-
graphical interface


Compute servers


-
Unix workstations


-
parallel computers


-
global condor pool

2000

www.nanoHUB.org

NCN

8

running applications with PUNCH

1994

AT&T

grant

2002

NCN

>7M hits (1994
-

2002)

www.nanoHUB.org

NCN

9

Network for Computational Nanotechnology

NCN

UIUC

NU

UTEP

Norfolk
State

Purdue

Berkeley


CN has a vision to pioneer the


development of nanotechnology
from science to manufacturing through
innovative theory, exploratory
simulation, and novel
cyberinfrastructure
.

N

‘an infrastructure and research network’

www.ncn.purdue.edu

www.nanoHUB.org

NCN

10

NCN Mission

NCN

UIUC

NU

UTEP

Norfolk

Purdue

Berkeley

1)
to connect computational
experts with experimentalist,
educators, and students


2)
to bridge disciplines and
promote collaboration


3)
to support CSE


4)
to disseminate knowledge and
services


5)
to
enable

research and
education

“cyberinfrastructure”

www.nanoHUB.org

NCN

11

NCN Outcomes

NCN

UIUC

NU

UTEP

Norfolk

Purdue

Berkeley

1)
Advances in nanoscience and its
transition to nanotechnology


2)
Pervasive, critical, and effective use
simulation in nanotechnology
research and education


3)
Advances in CSE


4)
Creation of a major, electronic
resource for nanotechnology


5)
Dissemination of technology and best
practices to other communities.

www.nanoHUB.org

NCN

12

Rappture = Rapid Application Infrastructure

Scientist


Created by NCN in Nov 2004


Works with your favorite


programming language



Open source


Online at http://rappture.org

Rappture

=

Simulation

Code

www.nanoHUB.org

NCN

13

The Rappture approach


standardizes interfaces



improves usability and speeds program debugging



complete record of each simulation



a strategy to develop high quality software quickly



and longer term, to assemble ambitious workflows

www.nanoHUB.org

NCN

14

Maxwell’s

Daemon

Physical Machine

Virtual Machine

middleware system architecture

Content

Database

Rendering Farm

nanoHUB cluster

Violin

nanowire job

nanowire job

nanowire job

1011

0101

1001

nanoVIS

www.nanoHUB.org

NCN

15

online simulation

more than 80 tools online


more that 100 in development

www.nanoHUB.org

NCN

16

nanoHUB tool page

user statistics

reviews and
citations

getting started

how to cite

launch!

www.nanoHUB.org

NCN

17

NCN’s software strategy

1)
facilitate the sharing of SW tools emerging from research


2)
disseminate high
-
quality simulation codes


3)
develop specialized tools for experimentalists and educators


4)
promote the intelligent, critical use of simulation



www.nanoHUB.org

NCN

18

more than simulation

tutorials and seminars

research seminars

learning modules

+
online meetings, Q and A,
reviews, SW development
tools, statistics, etc.…

online courses

www.nanoHUB.org

NCN

19

MIT
OpenCourseWare

“A free and open educational resource
-

for educators, students, and
self
-
learners around the world.” All 1800 MIT courses are now online.

www.nanoHUB.org

NCN

20

nanoHUB usage

www.nanoHUB.org

>65,000 users/year

www.nanoHUB.org

NCN

21

usage

www.nanoHUB.org

NCN

22

users


position


graduate student:


55%

undergrad student:

18%

pre
-
college student:


1%

scientist / engineer:

13%

faculty:



13%


(April 2006)


age


18
-
25:



61%

26
-
35:



29%

36
-
45:





7%

46
-
55:




2%

56 or older:



1%



(March 2006)


use nanoHUB for


research:


33%

education:


38%

both equally



28%

other:




1%



(November 2006)


technical interests


nanoelectronics:


46%

NEMS/nanofluidics: 9%

nanomedicine


11%

nanomaterials:


16%

nanophotonics:




8%



(April 2006)

www.nanoHUB.org

NCN

23

HUBzero.org



Linux/Apache/MySQL/PHP



LDAP authentication



Joomla Content Mgmt



Hub website components


-
tool development framework


-
web publishing systems


-
statistics collection / analysis


-
online meetings


-
topic pages


-
recommendation engine


-
Questions and Answers


-
incentive system


-
citations and DOO




Maxwell’s Daemon



Rappture Toolkit

www.nanoHUB.org

NCN

24

New Hubs Online

IndianaCTSI.org



Anantha Shekhar, IU

School of Medicine, Connie Weaver at Purdue

accelerating clinical and translational research in healthcare

online since 10/1/2007

thermalHUB.org



Tim Fisher, ME at Purdue

heat transfer

online since 12/6/2007

pharmaHUB.org



Rex Reklaitis, CE at Purdue

pharmaceutical product development and manufacturing

online since 12/11/2007

GlobalHUB.org



Dan Hirleman, ME at Purdue

global engineering education

online since 12/17/2007

nanohub.org



Mark Lundstrom, ECE at Purdue

the granddaddy of all hubs focused on nanotechnology

online since 2002

www.nanoHUB.org

NCN

25

impact

www.nanoHUB.org

NCN

26

Supriyo Datta

Supriyo Datta

Concepts in Quantum Transport

From Atom to Transistor

Fundamentals of Nanoelectronics

Electronics from the Bottom Up


9,999 nanoHUB users last year


‘datta’ is
the

most popular search term
on the nanoHUB

www.nanoHUB.org

NCN

27

M. Ashraf Alam

photo of you

Problem:

Approach:

Results:

For medium scale integration of carbon
Nanonet transistors for flexible electronics,
the contamination of metallic tubes makes
making large circuits difficult.

Develop fundamental understanding of
percolative transport so that the
threshold of percolation can be tuned for
specific circuits.

Theory of asymmetric percolation in
heterogenous system that allows development
of ~100 transistors integrated circuits on flexible
substrates.

Muhammad A. Alam

www.nanoHUB.org

NCN

28

Effect of metallic CNTs

www.nanoHUB.org

NCN

29

Theory

Striping: cutting the tubes for on
-
off ratio

expt

Qing Cao, et al., “Medium Scale Carbon Nanotube Thin Film Integrated

Circuits on Flexible Plastic Substrates,” to appear in
Nature,

2008

www.nanoHUB.org

NCN

30

IEEE EDL

Feb. 2007


Connection to NCN / nanoHUB

“The
finite
-
size percolation

model was
used to calculate the I
D
-
V
G

characteristics for
NanoNET transistor

with
channel length of 2 um

…”


promotes diffusion of knowledge


encourages collaboration


increases the impact of the work

www.nanoHUB.org

NCN

31

Network for Computational Nanotechnology

Problem:

Approach:

Results:

Atomic level structure of semiconductor
heterostructures controls their
electronic properties.

Molecular dynamics with interatomic
potentials derived from first principles

Tight binding for electronic structure

Size can be used to control strain in nanoscale heterostructures

Alejandro Strachan

www.nanoHUB.org

NCN

32

Strain relaxation in Si/Ge/Si nanobars

Si

Si

Ge

Bar width (W)

height

Simulations show that increasing
the bar height or decreasing its
width reduces transverse strain in
Ge section


Atomistic prediction in good agreement
with experiments



Theory can be used to optimize
material
in silico

before fabrication

www.nanoHUB.org

NCN

33

NCN / nanoHUB.org

nanoMATERIALS simulation toolkit: general purpose MD simulations

Input parameters

Interactive output:

molecular structure and graphs

nanoMATERIALS tutorial:
https://www.nanohub.org/resources/2322



Lecture series on MD:
https://www.nanohub.org/resources/3675


Materials Modeling and Simulation class (Fall 2008)

www.nanoHUB.org

NCN

34

Arvind Raman

photo of you

Problem:

To mathematically simulate the motion of nanoscale
Atomic Force Microscope probe tips scanning over
organic and inorganic samples

Approach:

Couple vibrating cantilever eigenmodes to realistic tip
-
sample interaction force models that include van der
Waals, electrostatic, repulsive interactions. Use special
integration routines to improve simulation speed and
accurately integrate across high force gradients.

Results:

Resonance enhancement in liquids for improved material contrast

Arvind Raman

www.nanoHUB.org

NCN

35

Resonance enhancement of harmonics in liquids

Higher harmonics of tip motions in buffer solutions for
the imaging of soft biological samples have been
simulated. Some harmonics are enhanced due to the
second eigenmode resonance. This is a generic
phenomenon in liquids for soft cantilevers used for
AFM imaging of biological samples.

Simulations predcted that if the images of these
resonance enhanced harmonics were mapped
across a sample, then significant improvement in
contrast of material properties is obtained
(proportional to local elasticity). Experiments
validated the predictions.

2T
0
T
Time
0
0
20
-
20
2
-
2
Slow Timescale
Fast Timescale
AFM Cantilever
Purple membrane (PM)
Mica substrate
(a)
Tip
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
DFT (dB)
Frequency
0
10
f
d
20
f
d
(c)
50dB
2T
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
0
T
Amplitude (nm)
Time
(b)
10nm
2T
0
T
Time
0
0
20
-
20
2
-
2
Slow Timescale
Fast Timescale
AFM Cantilever
Purple membrane (PM)
Mica substrate
(a)
Tip
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
DFT (dB)
Frequency
0
10
f
d
20
f
d
(c)
50dB
2T
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
0
T
Amplitude (nm)
Time
(b)
10nm
2T
0
T
Time
0
0
20
-
20
2
-
2
2T
0
T
Time
0
0
20
-
20
2
-
2
Slow Timescale
Fast Timescale
AFM Cantilever
Purple membrane (PM)
Mica substrate
(a)
Tip
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
DFT (dB)
Frequency
0
10
f
d
20
f
d
(c)
50dB
2T
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
0
T
Amplitude (nm)
Time
(b)
10nm
2T
0
T
Time
0
0
20
-
20
2
-
2
Slow Timescale
Fast Timescale
AFM Cantilever
Purple membrane (PM)
Mica substrate
(a)
Tip
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
DFT (dB)
Frequency
0
10
f
d
20
f
d
(c)
50dB
2T
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
0
T
Amplitude (nm)
Time
(b)
10nm
2T
0
T
Time
0
0
20
-
20
2
-
2
Slow Timescale
Fast Timescale
AFM Cantilever
Purple membrane (PM)
Mica substrate
(a)
Tip
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
DFT (dB)
Frequency
0
10
f
d
20
f
d
(c)
50dB
2T
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
0
T
Amplitude (nm)
Time
(b)
10nm
2T
0
T
Time
0
0
20
-
20
2
-
2
2T
0
T
Time
0
0
20
-
20
2
-
2
Slow Timescale
Fast Timescale
AFM Cantilever
Purple membrane (PM)
Mica substrate
(a)
Tip
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
DFT (dB)
Frequency
0
10
f
d
20
f
d
(c)
50dB
2T
Tapping on mica, SP=94%
Tapping on PM, SP=94%
Just before tapping
0
T
Amplitude (nm)
Time
(b)
10nm
Harmonic number

X. Xu, J. Melcher, R. Reifenberger, A. Raman,
“Resonance enhancement of cantilever higher
harmonics in liquids: enhancing compositional
contrast with gentle forces”, In preparation

www.nanoHUB.org

NCN

36

*
J. Melcher, S. Hu, A. Raman, “VEDA


a web based virtual environment for
dynamic Atomic Force Microscopy”, Invited article


Review of Scientific
Instruments, June 2008..

Review of Scientific
Instrumentation


A monthly journal devoted to scientific
instruments, apparatus, and techniques

June, 2008

NCN/nanoHUB.org


Increasing interest in CI to deliver virtual
instruments



Collaboration with DOE Molecular
Foundry to include realistic noise sources
into the current (deterministic) models in
VEDA



In addition to scientists and students
worldwide, VEDA is being used by major
US AFM/nano
-
instrumentation companies
such as Veeco, Agilent, and Asylum for
both training and research.

www.nanoHUB.org

NCN

37

cyberinfrastructure

“The conduct of science and engineering is changing and
evolving. This is due, in large part, to the expansion of
networked cyberinfrastructure.”


NSF Strategic Plan 2006
-
2011


www.nanoHUB.org

NCN

38

shared research facilities

Birck Nanotechnology Center, Purdue University


Courtesy HDR Architecture, Inc./Steve Hall


Hedrich Blessing

www.nanoHUB.org

NCN

39

“service
-
oriented science”

Distributed Computing

VIEWPOINT

Service
-
Oriented Science

Ian Foster


New information architectures enable new approaches to publishing and

Accessing valuable data and programs… as services….. Thus,
tools
formerly accessible only to the specialist can be made available to
all
;…Such service
-
oriented approaches to science are already being
applied successfully, in some cases at substantial scales….


6 MAY 2005 VOL 308 SCIENCE www.sciencemag.org

www.nanoHUB.org

NCN

40

lessons learned

NCN


it takes a dedicated core team with a
vision and something special to
share



need people who are ‘close to the
problem’ and ‘close to the solution’



people need to be doing the right
things



must be willing to adapt and evolve



IT and SW development is expensive
(so is assessment)

www.nanoHUB.org

NCN

41

NCN is a work in progress

science
drivers

network leadership

technology development

and support


refine and expand the SW collection



move from a ‘resource’ to a ‘community’



continue to enhance the infrastructure



expand coverage of nanotechnology



grow the user base



strengthen CSE engagement



develop a sustainability model


NCN

www.nanoHUB.org

NCN

42

NCN in the future

HUBzero.org

NCN

centers,
groups,
PI’s

universities

other
networks

other orgs

www.nanoHUB.org

NCN

43

cyberinfrastructure

“The conduct of science and engineering is changing and
evolving. This is due, in large part, to the expansion of
networked cyberinfrastructure.”


NSF Strategic Plan 2006
-
2011