What’s New in CISE:Status Report and

Reflections

Gregory R. AndrewsProfessor, Computer Science, Univ. of Arizona

Former Division Director, Computer and Network Systems, National Science Foundation

February 2005

2

Outline

• NSF context • The CISE Directorate• Programs, trends, and plans• Reflections and lessons learned

3

National Science Foundation (NSF)

• Created in 1950“to promote the progress of science; to

advance the national health, prosperity, and welfare; to secure the national defense; and for other purposes”

• Roles– Support basic research– Train the next generation– Educate the public– Advise the government on science policy

4

Federal Support for Research

• National Science Foundation (NSF)– Curiosity-driven basic research– Long term

• National Institutes of Health (NIH)– Use-inspired basic research– Long term

• Mission Agencies: DARPA, DOE, NASA, etc.– Applied research– Shorter term

5

Pasteur’s Quadrant

Basic research (Bohr) NSF 1950 … 2005

Use-inspired basic research (Pasteur)

NIH

Applied research (Edison)

Mission Agencies

6

NSF Organization

Directorate for BiologicalSciences

Directorate for Mathem atical& Physical Sciences

Directorate for Computer &Information Science and Engineering

Directorate for Social, Behvioral& Econom ic Sciences

Directorate for Education& Human Resources

Directorate for Geosciences

Directorate for Engineering Office of Polar Programs

Office of Integrative Activities

Office of the Director

National ScienceBoard

Administrative Offices

7

NSF Crosscutting Initiatives for 2005

• Biocomplexity in the Environment• Nanoscale Science and Engineering• Mathematical Sciences• Human and Social Dynamics

• Information Technology Research - ended as an initiative in 2004

8

Computer and Information Science and Engineering

(CISE)

• Created in 1985 (out of MPS)– Three research divisions– Two infrastructure divisions:

supercomputing and networking– Office of cross-disciplinary activities

• Minor reorganization in 1997– 5 Divisions: CCR, EIA, IIS, ACIR, ANIR

• Major reorganization in 2003– 4 Divisions: CCF, CNS, IIS, SCI

9

CISE Responsibilities

• Support basic research and education in computer and information science and engineering

• Support a shared cyberinfrastructure for all of science and engineering

10

New CISE Organization

C om puting andC om m unica tions

Foundations(C C F)

C om puter andN etw orkSystem s

(C N S)

In form ation andIn te lligentSystem s

(IIS )

SharedC yberin frastructure

(SC I)

O fficeof the

AssistantD irector

11

Key Concept: Cluster

• Comprehensive activity in a coherent area of research and education

• Team of program officers and staff working closely with the community

• Initially: group of existing programs• By end of FY05: one program per cluster

12

Computing and Communication

Foundations (CCF)• Formal and Mathematical Foundations

– Computer science theory; numerical computing; computational algebra and geometry; signal processing and communication

• Foundations of Computing Processes and Artifacts– Software engineering; software tools for HPC; programming

languages; compilers; computer architecture; graphics and visualization

• Emerging Models for Technology and Computation– Computational biology; quantum computing; nano-scale

computing; biologically inspired computing

13

CCF Competitions

• FY 2004– Responsible for about 2030 proposals– Heavy mortgages and NSF-wide commitments– Decent success rates for CAREER (15%) but terrible

success rates for clusters (6%)

• FY 2005 – Theoretical Foundations: January 2005– Emerging Models for Technology and Computation:

February 2005– Foundations of Computing Processes and Artifacts: May

2005 with awards in fall from FY 2006 budget

• FY 2006 and 2007– Possibly no competitions in FY 2006– Fall deadlines for all three clusters in FY 2007

14

Computer and Network Systems (CNS)

• Computer Systems– Distributed systems; embedded and hybrid systems; next-

generation software; parallel systems

• Network Systems– Networking research broadly defined plus focus areas in

programmable wireless networks and networks of sensor systems

• Computing Research Infrastructure– Research infrastructure; minority institutional

infrastructure; research resources

• Education and Workforce– Curriculum development/educational innovation; IT

workforce; special projects; cross-directorate activities (e.g., REU sites)

15

CNS Competitions

• FY 2004– Responsible for about 2035 proposals– Good success rates for CAREER and infrastructure (30%)– Fair success rates for research programs (18-20%)

• FY 2005 — One solicitation per cluster– Computer Systems: November 2004– Network Systems: January 2005– Computing Research Infrastructure: July 2005– Education and Workforce: Education with research

programs; workforce subsumed by Broadening Participation emphasis area

• FY 2006– Same deadlines as in FY 2005, but Networking in December

16

Information and Intelligent Systems (IIS)

• Systems in Context– Human computer interaction; educational technology;

robotics; computer-supported cooperative work; digital government

• Understanding, Inference, and Data– Databases; artificial intelligence; text, image, speech, and

video analysis; information retrieval; knowledge systems

• Science & Engineering Informatics/Information Integration– Bioinformatics; geoinformatics; cognitive neuroscience;

data-driven science

17

IIS Competitions

• FY 2004– Responsible for about 2590 proposals– Success rates 17% CAREER, 6% regular.

• FY 2005– Raise success rate of 2004 to 12-15%– Science & Engineering Informatics/Information Integration

and Universal Access: December 2004– Data, Inference, and Understanding and Systems in

Context: May 2005 with awards in fall from FY 2006 funds

• FY 2006– Same deadlines as in FY 2005

18

Shared Cyberinfrastructure (SCI)

Cyberinfrastructure: computational engines, storage, networking, data, sensors, software, and services to support advances in science and engineering

• Infrastructure Deployment– Planning, construction, commissioning, and

operations

• Infrastructure Development– Creating, testing, and hardening next-generation

deployed systems

19

History of NSF CI Investments

PACI NPACI and Alliance

Terascale TCS, DTF, ETF

Cyberinfrastructure

Supercomputer Centers

1985 1990 1995 2000 2005 2010

| | | | | |

SDSC, NCSA, PSC, CTC

Prior Computing Investments

NSF Networking

NMINSF Middleware Initiative

ITRInformation Technology Research

20

SCI Competitions

• FY 2004– Core funding for the PACI centers and expansion of the

Extensible Terascale Facility (ETF)– NSF Middleware Initiative (NMI): 141 proposals; 20 awards– International Network Connections: recently decided

• FY 2005– Continuing support for PACI and ETF– Cyberinfrastructure Teaching, Education, Advancement,

and Mentoring (CI-TEAM): Spring 2005– NSF Middleware Initiative (NMI): Spring 2005– Leveraging and coordinating shared and domain-specific

cyberinfrastructure with other agencies and directorates

21

Key Concept: Emphasis Area

• Focused area of research that cuts across clusters and divisions

• Addresses a scientific and/or national priority

• Has program announcement and funds

22

FY 2004 Emphasis Areas

• Cyber Trust– Develop computing systems that operate securely

and protect sensitive information– Received 488 proposals; made 50 awards; got $5M

in co-funding from DARPA• Information Integration

– Integrate and “mine” large data repositories to support data-driven science

– Received 238 proposals; made 33 awards• Science of Design

– Develop a body of theoretical and empirical knowledge to facilitate creation of a science of software design

– Received 182 proposals; made 24 awards

23

FY 2005 Emphasis Areas

• Information Integration: December 2004• Cyber Trust: February 2005• Science of Design: Spring 2005

• Broadening Participation: June 2005– Support alliances and projects that have the

potential significantly to increase the number of underrepresented students achieving college and graduate degrees

• Probably one more on High-End Computing

24

NSF and CISE Annual Budgets ($M)Actuals

0

1000

2000

3000

4000

5000

6000

7000

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Req

FY

NS

F B

ud

get

($M

)

0

100

200

300

400

500

600

700

CIS

E B

ud

get

($M

)

NSF

CISE

25

Funding Rate for Competitive Awards in CISE

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Nu

mb

er o

f P

rop

osa

ls a

nd

Aw

ard

s

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Fu

nd

ing

Rat

e

Competitive Proposal Actions Competitive Awards Funding Rate

26

NSF and CISE Funding Rate Trends

0%5%

10%

15%20%25%30%

35%40%

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

FY

Fu

nd

ing

Rat

es

NSF

CISE

27

CISE FY 2005 BudgetRequest ($M)

CISE Budget Lines FY 2005

CCF $91.41

CNS $132.39

IIS $92.54

SCI $123.60

ITR (cross-CISE) $178.11

CISE Total $618.05

28

Reflections on Being at NSF

• My Duties– Represent CISE in NSF and Interagency settings– Help set CISE directions and policies– Manage the development and execution of CNS

programs and budget– Manage CNS staff: scientific and administrative

• Program Officer Duties– Represent their discipline within CISE and NSF– Interact with community to get input and provide

advice– Help define directions for their area of research– Manage competitions: core and cross-disciplinary

29

General Observations

• Science Policy and Government Agencies– There are lots of smart, hardworking government

employees– Each agency, including Congress, has its own point of view

and its own agenda– Budget size matters

• National Science Foundation– Widely respected throughout government, for good reason– There is an institutional ethic to provide service to the

scientific community– It’s lots more fun when the budget is rising (2003) than

when it is falling (2004 and 2005)

30

General Observations II

• CISE Directorate– The importance of CISE is recognized within NSF– Funding decisions truly are guided by NSF’s dual roles:

• Supporting good science • Training the next generation—throughout the

country– The rapid increase in proposals has put CISE under

tremendous pressure• The scientific staff is overworked• The peer review process is at the breaking point

– Despite the above, panels and program officers are making good recommendations

– However, few projects are adequately funded and lots of really good work is not getting funded

31

Advice

• Attributes of winning proposals– Address important problem and have novel idea(s)– Well written project description, good technical depth,

know the related work– Address broader impacts and describe (own) prior

work — and read the proposal submission instructions!

• Interact with program officers– Get feedback on proposals, ask for advice, provide

input– Volunteer to be a reviewer

• Consider working at NSF at some point

32

Conclusion

• NSF’s role is fundamental to all areas of society — the most basic future investment

• Computer science and related disciplines are hugely important in their own right and essential to advancement in all areas of S&E

• NSF and our field are facing unprecedented pressures that can only be overcome by concerted, cooperative action

33

Further Information

• CISE Web site: www.cise.nsf.gov

• Computing Research News bimonthly columns: www.cra.org

• Contact: greg@cs.arizona.edu