11/30/11
nanoHUB.org powered by HUBzero® –
A Platform for Collaborative Research
with Quantifiable Impact on Research and Education
Gerhard Klimeck
Director Network for Computational Nanotechnology
gekco@purdue.edu
Gerhard Klimeck – Year 2002
Principal Scientist at NASA JPL:
•  Built Nanoelectronic Modeling Tool Suite
–  Multi-million atom electronic structure, Quantum transport in nanoelectronics
–  Designs of quantum Dot Detectors , Resonant Tunneling Diodes, etc.
–  Design tool with user interface
–  First Beowulfs / clusters built there in 1997/1998
•  BUT:
–  No means of wide dissemination of code, access, HPC resources
–  Portal technologies do not support interactive simulations
Group Supervisor of HPC Group:
•  Manage research collaborations
–  Internal groups
–  Other National Labs
–  University Groups.
•  BUT:
–  No means of fostering collaboration, rapid software exchange,
only one-on-one interactions
172 countries
Mark Lundstrom – Year 2002:
=> PUNCH to the next level
Problem Statement:
•  Modeling and simulation is underutilized in nanoscience and
nanotechnology
–  Software not shared, not accessible, not usable
–  Collaborations and knowledge transfer limited to personto-person process
–  Workforce development weak
–  Industry & experimental impact small
nanoHUB Vision in 2002:
•  Advance Nanoscience to Nanotechnology
–  Enable sharing models, access, and cycles of tools
–  Impact research and education
–  Impact industry
–  Build a operational infrastructure
Seen PUNCH – Purdue Network Computing HUB ☺
NCN vision 2002
accelerate the
transformation of nanoscience to nanotechnology
through simulation
1
11/30/11
NCN vision 2002
1965
Relative Manufacturing Cost per Component
Gordon Moore
enable
new modes of discovery, innovation, learning, and engagement
that accelerate the
transformation of nanoscience to nanotechnology
through simulation
tightly linked to experimental research and education
http://www.intel.com/technology/mooreslaw
Number of Components per Integrated Circuit
Berkeley
Simulation Program with Integrated Circuit Emphasis.
Intel in 2009
Device Size:
Tens of nanometers
from: Larry Nagel, BCTM ‘96
• Started as a class project
Stanford SUPREM
• Developed as a teaching tool
RobertChau(Intel), 2004
• Quality control: pass Pederson
Ronald
A. Rohrer
http://www.intel.com/technology/mooreslaw/index.htm
Device Integration:
>2 Billion
Berkeley SPICE
htt
p:/
/w
ww
.om
eg
a-
Laurence
W. Nagel
en
ter
pr
ise
s.n
et/
Donald O.
Pederson
• Dissemination:
Public domain code
Pederson carried tapes along
Students took it along
to industry and academia
Released 1972
2
11/30/11
Stanford
Stanford University PRocEss Modeling
Device Size
• Stanford wanted to mimic
Berkeley success
Birth of an Industy
P
Si roc
mu es
lat s
ion
Intel Capitalization:
$85B
• Combine various existing models
• Dissemination:
Total Industry:
Transistors
Public domain code
Community workshops
Students took it along
to industry and academia
it
rcu tion
i
C ula
m
Si
$280B
Years
na
str no-s
uc ca
tur le
es
What’s Next?
Nano Initiatives
Ele
ro
ct
nic
s
Device Size
Device Size
What’s Next?
na
str no-s
uc ca
tur le
es
Nano Initiatives
Ele
rials
Mate
na
of es
s
n tur
llio c
Bi stru
Years
Bi
Research
Photonics
Mec
o/
Me
han
di
cin
ics
e
nic
s
rials
Mate
Transistors
Transistors
Research
no
ro
ct
no
na
of es
s
n tur
llio c
Bi stru
Bi
Photonics
Mec
o/
Me
han
di
cin
ics
e
Years
3
11/30/11
:
Imagine
Goals - Impact Metrics
•  Services:
–  Modeling and Simulation Software
–  Seminars, tutorials, classes
•  Goals:
–  Knowledge transfer
•  Use in class rooms
–  Knowledge generation
•  Use in research
•  Use by experimentalists
–  Economic impact
•  Use in Industry
–  Professional Development /
Community building
E
t
lec
Mat e
Research
Bi
ni
ro
Pho
Mec
o/
Me
ha
di
:
Imagine
eway
t
ce Ga
n
e
i
c
l/S
Porta
Dream
ci
:
Imagine
No, Really!
What is Needed?
No, Really!
What is Needed?
ay
Gatew
ience
l / Sc am
ta
r
o
e
P
Dr
Technology
eway
t
ce Ga
n
e
i
Understanding
c
l/S
Porta
Dream
Stamina
Technology
Geek
User
CI
Ops
Understanding
Stamina
Virtual Org.
4
11/30/11
Mythbusting
Scientific Knowledge Transfer
User Perceptions / Beliefs:
•  Cannot use research software for
education
•  You cannot use someone else’s
code to conduct research
–  Experimentalists will not use
computational research codes
•  Codes are too hard to install
•  Codes get out of date
•  You cannot provide enough
compute cycles
Geek
User
CI
Ops
Virtual Org.
Mythbusting
Scientific Knowledge Transfer
CI Operation Perceptions / Beliefs:
•  Need one designated computer
scientist per application to port to
web and to support
=> $200k per application / year
•  A University cannot create and
serve a National Resource
•  There is no infrastructure that is
–
–
–
–
Secure
Serves users and developers
Affordable
Scalable
Mythbusting
Scientific Knowledge Transfer
Developer Perceptions / Beliefs:
•  No incentive to share working
code
•  Scientists must rewrite their
codes for web deployment
•  Graphical user interfaces cannot
be built by scientists / engineers
=> Scientists must hand-over their
code to someone else
=> Scientists disowned
=> Scientists will never use their
own code on the web platform
Macroscopic
dimensions
CI
Ops
Virtual Org.
ive
tic
llis
Ba
Drift /
Diffusion
Un
S
Virtual Org.
CI
Ops
Law of Equilibrium :
Non-Equilibrium
Quantum
ρ = exp (−(H
− μN) /kT )
Statistical
Mechanics
fus
Geek
User
Quantum Transport far from Equilibrium
Dif
User
Geek
ifi
ed
Atomic
dimensions
tum
n
ua
Σs
Q
Boltzmann
Transport
mo
de
μ1
Non-Equilibrium
Which
Green
Functions
Formalism?
l
D
SILICON
VG
S
Σ1
Σ2
D
INSULATOR
VD
Gerhard
Gerhard
Klimeck,
Klimeck
Supriyo Datta
I
μ2
H
VG
VD
I
5
11/30/11
NEMO
Nanoelectronic Modeling
OMEN Scaling to 221,400 Cores
Engineering at the Peta-Scale
g
Result:
in
• Highly efficient parallel
er
e
algorithm, stressing
!
n the
gi t the ale
most advanced
n
c
a today
resources E
available
S
-
ta
Pe
18 years of development
• Texas Instruments
• NASA JPL
Atomistic Device representation
• Purdue
• Deemed by many too
computationally intensive…
Impact
• Move from nano-science
to nanodevice engineering
in minutes
• Unprecedented insight
into atomistic device
simulation
Gerhard Klimeck
Gerhard Klimeck
Transistors at the end of the roadmap
MIT InAs HEMT
Goals:
• High ON current (high drive)
• Low OFF current (low loss)
• Fast switching
Experimental Approaches:
• Reduce device size
• Strain engineering
• New materials (III-V materials)
Experimental Problem:
• OFF Current too high
Gate
InGaAs
InAs
InGaAs
Drain
Source
InAlAs
InAlAs
?
Simulation Domain
Experiments:
D.H. Kim et. al, IEDM 07, EDL 08
Gerhard Klimeck
Gerhard Klimeck
6
11/30/11
Include Rounded Corners and Gate Tunneling
=> Match Experimental Data
Lg = 30nm
gy
lo
tro
ns
io
at
Gerhard Klimeck
Electronic Structure &
Electron Quantum
Transport Simulations
Drain
Source
e
ul
/ M t! sim
ts r
Lg = n40nm
e apa cale
Gate
InAlAs
S
em s
ur vice nos
InGaAs
InAs
ea e Na
InGaAs
M ak d by
rd e
InAlAs
da br logy Lg = 50nm
Simulation Domain an
•  Match experimental data
ro
St
et
•  Provide metrology
M
w:
•  Begin to guide experiment
Ne
•  Share code with experimental
=> group
Gerhard Klimeck
Over 220 tools online!
NEMO & OMEN
Runs on fastest computer
in the world
Runs 6 tools in nanoHUB
>6,700 users
>100k sims
7
11/30/11
Over 2,900 Resources!
>220 tools
>55 courses
Over 2,900 Resources!
>220 tools
>55 courses
2,200 seminars and
teaching materials
nanoHUB Video
2,200 seminars and
teaching materials
Nano App Store
188,000 users worldwide
As much traffic as www.purdue.edu
Users at all Top 50 US Engr Schools
19% of all .edu domains
172 countries
Users in
Sept 2010
8
11/30/11
Sociology
Nano App Store
188,000 users worldwide
As much traffic as www.purdue.edu
Users at all Top 50 US Engr Schools
19% of all .edu domains
How do Users Behave?
•  Questions:
–
–
–
–
–
How many students in the class?
Which tools?
Intensity of use
Sustained use
Percentage of service: Education vs. Research use
•  Some Statistics
–  8,600 users ran 345,000 simulations Academic Year 2009/2010
–  116 classes / 97 institutions in Academic Year 2009/2010
•  Info Obtained from self-registration, manual follow-up
–  575 citations in the literature
•  Info obtained from Google Scholar and manual analysis
172 countries
Formal Education vs. Research
Users
The nanoHUB Matrix
Time (days)
Each dot represents simulation activity on a particular day
The color of the dot indicates a particular tool
We will look backwards into history
for each user in the past 12 months
And plot ALL their activities
9
11/30/11
AY 09/10:Education
Formal
116 Courses,
97 institutions,
~2,100 students
95% outside NCN
vs. Research
Resource Requirements
Simulations vs .CPU Consumption
Myth Busted!
Proof of use in EDUCATION!
Knowledge Transfer out of Research
Voluntary / Viral Use
Resource Requirements
Simulations vs .CPU Consumption
Resource Requirements
Simulations vs .CPU Consumption
10
11/30/11
Resource Requirements
Simulations vs .CPU Consumption
Research:
• Avg:
200 runs, 8 hours CPU
• Top:
10,000 runs, 10,000 hours
• Education:
• Avg:
20 runs, 5 minutes
• Top:
400 runs, 20 hours CPU
Mythbusting
Scientific Knowledge Transfer
User Perceptions / Beliefs:
•  Cannot use research software for
education
•  You cannot use someone else’s
code to conduct research
–  Experimentalists will not use
computational research codes
•  Codes are too hard to install
•  Codes get out of date
•  You cannot provide enough
compute cycles
Geek
User
CI
Ops
oo
No f o
t j f us
us e
t c in
om E
pu xpe
ta rim
tio e
na nt
l t al W
he
or ork
y! !
Virtual Org.
Pr
! N
CH NC S
R
R
n
EA no HE
S
T
RE 7% O
in s, 7 by
e
e r
us tho y us
f
u
r
f o a ta
oo 200 lun
r
P 1, vo
r
f
ve o
O roof
P
43
43
44
44
11
11/30/11
2008 Data
Focus on Research
User Myth: No Good Research!
Academy of
Engineering
Member
Faculty
member
3 years
after PhD
213 Citations
575 nanoHUB citations
>3,200 secondary citations
h-index: 27
Research 213 80%
Res/Edu 9 3%
Education 12 5%
Cyberinfr 31 12%
45
Focus on Schred
80 Citations
Who is using this?
Experimentalist
Akiko Ohata
IMEP Minatec, France
Electrical characteristics related
to silicon film thickness in
advanced FD SOI–MOSFETs
Research 79
Cyberinfr 1
Ultra-thin fully-depleted SOI
MOSFETs: Special charge
properties and coupling effects
Simulation Runs
Web Visits
Publication
Source Code Download
12
11/30/11
Who is using this?
Other Users
Theorist
Enrico Sangiorgi
Scaling the High-Performance
Double-Gate SOI MOSFET down to
32 nm Technology Node with SiO/
sub2/-based Gate Stacks
University of Bologna, Italy
Analysis of Scaling Strategies for
Sub-30 nm Double-Gate SOI
N-MOSFETs
Simulation Runs
Web Visits
Publication
Other Users
Source Code Download
PADRE
Industrial Tool – Bell Labs
13
11/30/11
MOSFET: Running PADRE Simply
Impact of Reduced Tools
MOSFET:
2,715 Users,
38,000 jobs
MUGfet:
240 Users,
3,600 jobs
945 Users,
41,285 jobs
MOSCAP:
1,694 Users,
18,000 jobs
6,649 Users,
104,282 jobs
PN junction:
3,563 Users,
33,000 jobs
BJT:
557 Users,
3,000 jobs
Drift-Diffusion:
721 Users,
7,400 jobs
Importance of a good GUI
CNTbands
55
Same behavior across all similar converted tools
14
11/30/11
Balancing
Usability and Capability
Mythbusting
Scientific Knowledge Transfer
User Perceptions / Beliefs:
•  Cannot use research software for
education
•  You cannot use someone else’s
code to conduct research
–  Experimentalists will not use
computational research codes
•  Codes are too hard to install
•  Codes get out of date
•  You cannot provide enough
compute cycles
nanoHUB
Developer Perceptions / Beliefs:
•  No incentive to share working
code
•  Scientists must rewrite their
codes for web deployment
•  Graphical user interfaces cannot
be built by scientists / engineers
=> Scientists must hand-over their
code to someone else
=> Scientists disowned
=> Scientists will never use their
own code on the web platform
User
CI
Ops
Virtual Org.
iPhone / iPad
Mythbusting
Scientific Knowledge Transfer
Geek
Web-enabling Tools
2 years => 1 week
100 : 1 ratio
Geek
User
CI
Ops
Virtual Org.
Vendor
Scientist
Web Developer
15
11/30/11
Cyberinfrastructure for Running Tools
Building Interfaces and Data Management Systems Fast
Linux/Apache/MySQL/PHP
Rappture
=
Physical Machine
Content
Database
Maxwell’s
Daemon
Middleware
Virtual Machine
0101
1011
1001
nanowire job
Simulation
Code
tool session cluster
Rendering Farm
nanoVIS
Scientist
Developer Collaboration Network
White dots:
outside NCN
•  Rapid Application Infrastructure
•  Created by NCN in Nov 2004
•  Open Source (rappture.org)
•  Create standard desktop apps
•  Works with your favorite
programming language
Each dot is a tool
Links are people
Network Nodes Serving Users
Each dot is a site
Links are users
18,600
348
2,122
16
11/30/11
Next Generation Publications
Research Incentives
Tool Usage reading papers
Developer Collaboration Impact
Dragica Vasileska
Orange dots:
Site Leads
White dots: Developers
17 tools
11,570 users
123 citations
Next Generation Faculty:
Usage at SIUC
nanoHUB on iTunes U
Shaikh Ahmed
6,183 users
8 tools
Get images from annual report
Post Doc
at Purdue
Faculty at
SIUC
•  Infused nanoHUB into
existing classes
•  Built a new nanoelectronics
curriculum
•  Used nanoHUB for research
Recently Dr. Ahmed was promoted to tenured Associate
Professor. I would like to emphasize that Dr. Ahmed's use
of nanoHUB in education and research, which earned him
national and international visibility, did play a significant
positive role in his early promotion case.
Nov 2009 start
350 content items today
55,000 downloads
~10,000 downloads/month
Glafkos Galanos
Chair, Dept. of Electr. and Comp. Eng, SIUC
17
11/30/11
Wikipedia Contributions
German
Punjabi
Italian
16 animations deployed Jan 2010 on ~30 pages
Brings 2,200 visitors for 4,000 visits monthly
Mythbusting
Scientific Knowledge Transfer
–
–
–
–
Secure
Serves users and developers
Affordable
Scalable
Developer Perceptions / Beliefs:
•  No incentive to share working
code
•  Scientists must rewrite their
codes for web deployment
•  Graphical user interfaces cannot
be built by scientists / engineers
=> Scientists must hand-over their
code to someone else
=> Scientists disowned
=> Scientists will never use their
own code on the web platform
Geek
User
CI
Ops
Virtual Org.
Developer Activities
Geek
User
CI
Ops
Virtual Org.
CI Operation Perceptions / Beliefs:
•  Need one designated computer
scientist per application to port to
web and to support
=> $200k per application / year
•  A University cannot create and
serve a National Resource
•  There is no infrastructure that is
Mythbusting
Scientific Knowledge Transfer
18
11/30/11
Software Updates and Deployment
Cyberinfrastructure for Developing Tools
Tool
Developer
Cyberinfrastructure for Developing Tools
Web-based
Publishing
System
Registered
Registered
Created
Created
Uploaded
Uploaded
Installed
Installed
Approved
Approved
Published
Published
End User
>200 S/W projects
>300 developers
Managed by ONE person!
Tool Lifecycle Management
Good news!
In just
5 years, nanoHUB.org hosts
190 new tools plus
1000 version updates
Bad news!
In 1 year, 1/3 tools
are new or updated
In 2 years, 2/3 tools
are new or updated
– MUST MANAGE THIS
Good news!
In just 5 years, nanoHUB.org hosts
190 new tools plus
1000 version updates
Tool Lifecycle Management
19
11/30/11
Tool Support Technology Problems
Server-side tools have inherent
lifecycle management
Problem:
•  All support tickets were handled by central nanoHUB team
•  Tools owners decoupled unless contacted by nanoHUB staff
Approach
•  Develop triage system: ticket -> question -> wish
Current versions served by default
– PROBLEM SOLVED
rom
ar
Ye
09
20
sf
Gerhard Klimeck
Tool Support Technology Solution
Developers see
Problem:
tickets
on their
central
nanoHUB team
•  All support tickets were handled by
my HUB by
page
•  Tools owners decoupled unless contacted
nanoHUB staff
Approach
•  Develop triage system: ticket -> question -> wish
de
Sli
78
Virtual Economy Assessment
•  Introduced Questions and Answers Forum 2 years ago.
•  After one year, lots of questions, few answers
•  Introduced a virtual economy
»  Incentive point system
»  Value questions and answers
»  Royalty payments
•  Points can be spent to
»  Incentivize Answers
»  T-Shirts
(top choice on our earlier poll)
»  Future:
more storage, faster computing
m
2
Gerhard Klimeck
fro
s
Strategy for NCN and community supported
de tools
Sli
ar
Ye
09
20
•  Conducted a study comparing
pre- and post-incentive Q&A
forum
»  3x to 8x increase in Q&A activity
•  Will roll-out more incentives: tool usage royalties, reviews
es
79
m
fro
ar
Ye
09
20
d
Sli
20
11/30/11
User Contributions: the Challenge
• User contributions are key to scalability and
sustainability
Types of Contributions: All are Needed
• Tools/Educational Content:
Primarily contributed by select few members
» Most important source of knowledge transfer
» High-cost contribution, with high reward – contributors
publish “badge” on website and receive great
recognition in the academic community
• HUBs must have mechanisms to encourage
participation and [high quality] contributions
Anya Savikhin
• Need to understand the motivation behind behavior
to create proper mechanisms
» Are there any special characteristics of this
sample of users? Scientific community of users
who are not usually anonymous – quite different
from most online interactions
» Do the same incentives work for this sub-group as
for leisure communities (studied more often - e.g.,
Yan Chen et al. study on MovieLens)
• Rating of Content: Content rarely rated
» Low cost and low benefit, but requisite for selfsustainability – to build recommendation systems and
push most important content forward without relying on
site creator to rank content
» Many people must rate in order to develop system
(relying on a few ratings will be detrimental to a system)
• Q&A Forum: Questions rarely posted/answered
» Low cost and low benefit, but may build “community”
and indirectly increase rating of content (indirect
reciprocity)
Approach: Field Experiments
• Use concepts from behavioral economics to
provide effective incentive mechanisms, test
with experimental methods
• First study: On nanoHUB, we randomized users
to receive different types of incentives and
messaging and study their effects on (basic
survey) participation (Savikhin and Klimeck ,
2011)
• Users either receive 250 ‘nano
points’ (redeemable for gifts on the site), or
‘private benefit messaging’ or ‘social benefit
messaging’ to visit our site and fill out a survey
Anya Savikhin
Findings & Impact of First Study
• Findings
Anya Savikhin
» Users more likely to respond to a ‘private benefit’ message than a
‘social benefit’ message (unlike findings from leisure sites)
» Users twice as likely to respond when there is a ‘nano point’ offer
» In survey, we find most users contribute high-cost content (e.g.,
tools) to gain prestige
• Impact for Design of Incentive Mechanisms
» Recognition systems enhance private benefits and should be used
more; social feedback may not be as effective
» Employing recognition systems (badges, levels) for low cost, low
benefit tasks will be effective as it will increase private benefit –
lack of anonymity is an advantage here
» Employing ‘nano points’ for low cost, low benefit tasks will be
effective – but need more experiments to determine amounts/
sustainability
21
m
fro
es
6
id
Work our way up
>1 hour
N CPUs
2009
Parallel Batch
>1 hour
N CPUs
Serial Batch
Serial Batch
~10 minutes
1 CPU
~10 minutes
1 CPU
Fully Interactive
Fully Interactive
Seconds
Many jobs / CPU
Start here
Issue raised at the last site visit:
Can you provide this kind of ambitious
service?
Actual
Numbers
Applications
Parallel Batch
Work our way up
00
online simulations and more
Sl
Sl
id
Applications
Target Users and Requirements
r2
nanoHUB.org
Issue raised at the 2005 site visit:
Can you provide this kind of ambitious service?
es
fro
m
Ye
a
online simulations and more
Ye
a
00
Target Users and Requirements
r2
nanoHUB.org
6
11/30/11
Start here
Seconds
Many jobs / CPU
Lots of users in this space
No Exponential Fall-Off!
Network for Computational Nanotechnology
Network for Computational Nanotechnology
Mythbusting
Scientific Knowledge Transfer
HUB HPC
•
•
•
•
•
•
•
User Performance focused
Instant Access
Supercomputer access as needed
Large Cloud access as needed
Integrated data management
Visualize and compare results
Secure access
CI Operation Perceptions / Beliefs:
•  Need one designated computer
scientist per application to port to
web and to support
=> $200k per application / year
•  A University cannot create and
serve a National Resource
•  There is no infrastructure that is
–
–
–
–
Secure
Serves users and developers
Affordable
Scalable
Geek
User
CI
Ops
Virtual Org.
22
11/30/11
Mythbusting
Scientific Knowledge Transfer
Device Size
Geek
na
str no-s
uc ca
tur le
es
Ele
User
CI
Ops
Virtual Org.
ro
ct
nic
s
rials
Mate
Research
Transistors
User Perceptions / Beliefs:
•  Cannot use research software for
education
•  You cannot use someone else’s
code to conduct research
–  Experimentalists will not use
computational research codes
•  Codes are too hard to install
•  Codes get out of date
•  You cannot provide enough
compute cycles
Industry Impact
no
a
fn s
o
ns ture
li lio truc
B s
Bi
Photonics
Mec
o/
Me
han
di
cin
ics
e
Years
Papers by Industrial Authors
8.7%
OMEN
Mark Stettler, PhD.
Mgr. TCAD, Oregon
Publication
s
Runs on fastest computer
in the world
Electronics
Perfect scaling to
222,720 cores
• 46 papers total
• 41 papers in nano
Dmitri Nikonov, PhD.
Mgr. Strategic Research, CA
23
11/30/11
3 AFM Manufacturers:
New Partnership
•  Training / Virtual Instrument
•  Research
Virtual AFM
Mechanics
Materials
Mechanics
Fund tool development
I have been using VEDA …
… found it to be extremely useful. …
… enabled us to make better choices in designing new probes.
… used VEDA as a check on other calculations.
Roger Proksch
Asylum Research
Molecular Dynamics
(virtual surfaces)
Virtual Atomistic
Tip and Surface
VEDA
(Virtual AFM)
Goals - Impact Metrics
•  Services:
–  Modeling and Simulation Software
–  Seminars, tutorials, classes
•  Goals:
–  Knowledge transfer
•  Use in class rooms
–  Knowledge generation
•  Use in research
•  Use by experimentalists
–  Economic impact
•  Use in Industry
–  Professional Development /
Community building
c
Ele
ni
tro
What D
o User
Mat e
Research
Bi
s Say?
Pho
Mec
o/
Me
ha
di
ci
Top 75 words
•  159 notable quotes
•  527 content reviews
24
11/30/11
nanoHUB.org
The World’s Largest Nano User Facility
188,000 lecture users
830 lecturers
11,000 simulation users
300 sim developers
Top 75 words
•  159 notable quotes
•  527 content reviews
Users in 2010
172 countries
nanoHUB.org
Fully Operational Cloud for End Users
Systemic Use in Research:
•  Citation map shows
Social network of researchers
•  Enable other researchers to utilize
recent PhD thesis nano-modeling
software
•  Can measure tool impact
3
Any Science Gateway’s Dream
Why is it so hard?
Sc Any
ien
ce
Research
Cohorts of students behave similarly
Can measure class room sizes
Can measure tool impact
Time from research to classroom use
as small as 2 years!
g
y
An erin
e
gin
En
…
ge
an
Ch
3
Systemic Use in Education:
•
•
•
•
Impact:
•  Research: 719 citations in the literature
•  Education: 134 courses, 97 institutions (2010)
•  Collaboration: 300 simulation tool developers
220 simulation tools
..the..
...
wo
rld
25
11/30/11
Any Science Gateway’s Dream
There are worlds between…
7 Criteria
for Successful
Science Gateways
Research
1: Outstanding Science
2: Commitment to Dissemination
“Stuff the world wants”
“faculty that want to give it away”
46 faculty
Leveraged Research
$5.1M
+ 6 site leads
106 grad students
26
11/30/11
Next Generation Faculty:
Faculty Incentives
Usage at SIUC
Tool Usage reading papers
Dragica Vasileska
17 tools
7,835 users
115 citations
Shaikh Ahmed
6,183 users
8 tools
Get images from annual report
Post Doc
at Purdue
Faculty at
SIUC
•  Infused nanoHUB into
existing classes
•  Built a new
nanoelectronics
curriculum
•  Used nanoHUB
for research
Proof of Impact!
Great in Proposals!
Early Tenure Promotion
3: Technology for Dissemination
3: Technology for Dissemination
“simple and utterly dependable”
Less than 20 hours downtime
last year!
$1M/year
operation and
bridge building
27
11/30/11
4: Tech Transfer Processes
5: Understanding Stakeholders
“dedicated technical site leads”
Consultants
Geek
CI
Ops
Content Creation and Support
$2.2M
Significant
portion of budget
5: Understanding Stakeholders
User
6: Open Assessment / Incentives
“gather, understand, disseminate stats”
Geek
User
Geek CI
Ops
CI
Ops
Access,
Use,
Impact
User
Virtual Org.
28
11/30/11
7: Business Model
Research:
Sustained Academic Funding
Product Development:
Real Business Plans
What’s Next?
•  Start small:
–  A mirror HUB for South America
–  Hosted in Colombia
•  Think BIG!
Global Infrastructure Institute for
Engaged Science and Engineering
Hubs ‘R Us
hubzero.org
•  Feb 2007: 1 hub
•  Feb 2008: 5 hubs
•  Feb 2009: 8 hubs
•  Feb 2010: 21 hubs
•  Sept 2010: >30 hubs
HUBzero Consortium
Each hub has its own
funding stream
Outside institutions:
EPA, NYSTAR, Rice
2022 Vision:
Purdue recognized for driving US competitiveness and workforce
development by our ability to closely link advanced knowledge from
research into daily operations of industry
2022 Mission:
Use HUBzero to build Purdue preeminence in Presidential initiatives
•
» Nanotechnology .................... $1B/year, 3 presidents
» Manufacturing ....................... $0.5B, announced 6/26/2011
» Materials Genome Initiative ... $1B, announced 6/26/2011
» Healthcare Engineering
29
11/30/11
Global Infrastructure Institute for
Engaged Science and Engineering
Actions:
- multiple briefings to Aneesh Chopra, CTO, OSTP => manufacturingHUB
- briefing of Tom Kalil, OSTP => materialsHUB
- briefing DOE, Harriet Kung, Linda Horton => materialsHUB
- briefing ONR, Julie Christodoulou => materialsHUB
- briefing to Carol Tolbert, NASA
- briefings to AFRL – Air Force HUB prototype
2022 Mission:
Use HUBzero to build Purdue preeminence in Presidential initiatives
•
» Nanotechnology .................... $1B/year, 3 presidents
» Manufacturing ....................... $0.5B, announced 6/26/2011
» Materials Genome Initiative ... $1B, announced 6/26/2011
» Healthcare Engineering
Gl
ob
a
Ne l HU
tw B
or
k
A Federation of HUBs
It Takes a Village ...
US DOD and DOE
HUBs
Global HUB Partners
Colombia
... and Experience ...
Midwest Project for SME-OEM Use of Modeling and Simulation
OEMs: P&G, Lockheed Martin, John Deere, GE Energy
with small- to medium-size potential supply chain partners
Solution providers: Purdue, OSC, NCSA, NCMS
Improve competitiveness and innovation
HUBs
Industry HUBs:
Boeing, Lockheed,...
Success Criteria Identified with nanoHUB
1. Outstanding Science
2. Commitment and Incentives to Dissemination
3. HUB Technology
4. Tech Transfer Processes (dedicated staff)
5. Understanding Stakeholders
6. Open Assessment, Stats, Impact Metrics, etc.
7. Business Operational Model
... and serious
steps
30
11/30/11
NCN Vision for 2012 and Beyond
Jump back to simulation or data
Elevate computational science: make it reproducible
Gerhard Klimeck
Gerhard Klimeck
Explore validity, alternative scenarios
Write a new “paper”
metadata
copy
Gerhard Klimeck
Gerhard Klimeck
31
11/30/11
Write a new “paper”
Archive and explore data on nanoHUB
metadata
There will be many papers!
There will be lots of data!
All of it already has a meta data description!
How do we explore all this data?
Gerhard Klimeck
Gerhard Klimeck
NCN Vision: 2012
•  nanoHUB.org continues to grow
•  Partners with more nano centers and publishers
Klimeck
Research
Group
:
Imagine
»  Become THE central repository for nano knowledge
•  Elevate computational science: make it reproducible
»  Become THE central repository for nano data
•  Enables rapid exploration, understanding, and utilization
•  Revolutionize education, computational science, & publishing
Gerhard Klimeck
and
HUB
nano ro
e
z
HUB
Team
ay
l/
Porta
tew
ce Ga
Scien
m
Drea
32
11/30/11
Thank You!
Typical Dissemination Paths
"
" "
"
#
$
" 129
nanoHUB Technical Solution
"
'
"
(" $
Problems:
•  REALLY LONG stove pipe
• Web content: afterthought
usually stale
• Data shared by email
• Tools spread by hiring
%
Other Hubs
" "
29,707 users worldwide
"
#
$
" '
"
(" $
7,725 users worldwide
4,296 users worldwide
Problems:
•  REALLY LONG stove pipe
• Web content: afterthought
usually stale
• Data shared by email
• Tools spread by hiring
2,465 users worldwide
1,496 users worldwide
%
33