oct 25, 20051 program perspectives on quantum information michael foster national science foundation
TRANSCRIPT
Oct 25, 2005
1
Program Perspectives on Quantum Information
Michael Foster
National Science Foundation
Oct 25, 2005
2
Outline
• Computing & Communication Foundations
• QIS Bumper Stickers
• Quantum Computing
• Quantum Key Distribution
• Support agencies
• Where next?
Oct 25, 2005
3
Why Foundations?
-=
Oct 25, 2005
4
Foundations Everywhere
Infrastructure
Systems
Oct 25, 2005
5
Quantum InformationBumper Stickers
• Quantum computation– State superposition provides parallelism
• Quantum communication– No cloning theorem provides unforgability
• Quantum metrology– Entanglement provides consistent measurement
Oct 25, 2005
6
Quantum Computation
Oct 25, 2005
7
QIP and Moore’s Law
Year
1850 1950 20001900 2050
10-6
103
1
10-3
106
109
Babbage Engine
CMOS ICs
TX-2
ENIAC
Differential Analyzer
GeneralArchitecture
Lattice-GasArchitecture
Quantum Dots
Liquid NMR
Conve
ntio
nal C
ompu
ter R
oadm
ap
QC
Roa
dmap
MIPS
Oct 25, 2005
8
Power of Qubits
•Qubit = state of a quantum two-level system
1,1022 Continuum of states!
1 classical bit has two states: 0 and 11 qubit has “infinitely” many states!
•Physical realizations of qubits:•photon polarization•electron spin•nuclear spin•pair of electron states in a trapped ion/atom•magnetic flux state in a Josephson junction ring•Cooper pair number states, etc.
Oct 25, 2005
9
Power of Qubits
A classical 3-bit state: 001A quantum 3-qubit state:
111110101011100010001000
N qubits is worth 2n classical bits!•Entanglement
•Multiple qubits
101011100100
1001
1100
Not entangled
Entangled!
PhotonsOptical
Parametric Amplifier
OpticalParametric Amplifier
EPR photon pairEPR photon pair
or
NEVER
Oct 25, 2005
10
Power of Quantum Computation
•Quantum Parallelism
An exponential amount of computation has been achieved in the time it takes to compute the function on a single input!
2n values of F(x) all in one go!!
Oct 25, 2005
11
NP BQP?
Provably Q-hard algorithms
PSPACE
BQP
P
NP
Factorization (Shor)
Unsorted Search(Grover)
GraphIsomorphism?
Protein Folding?
Ajtai–Dwork Coding?
Quantum Complexity
Oct 25, 2005
12
Permutation Order-Finding (Chuang et al, ‘00)•Permutation is an operation that rearranges a set of objects
•Order r of a permutation applied to element y of a set is the minimum number of times must be applied to put y back in its original position
•Problem has wide range of applications (Cryptography)
QIP: An Example Algorithm
Oct 25, 2005
13
Classical versus Quantum
Classical approach:•Series of trials to find the x-th permutation x(y).
•Find equality. When a(y)= b(y) then ord() | a-b
•Number of trials needed increases exponentially with the number of bits representing y
Quantum approach:•Order is the period of a functionfy(x) = x(y)
•Quantum Fourier Transform allows us to find periods of all x(y) with one transform
•Exponential speedup-- Minimum number of steps proportional to bits in y
Oct 25, 2005
14
Classical Order-Finding
2
3
rYY
3
7
AA
ZZ
MM
AA
FF
2
26
15
Check Equality
Classical approach:
What a permutation really looks like:
Oct 25, 2005
15
Quantum Order-Finding
YY
3
7
AA
ZZ
MM
AA
FF
2
26
15
Quantum approach:
Equality
Large Fourier Components
Quantum Fourier
Transform
Checks Equality in
Parallel
Likely result
Oct 25, 2005
16
Quantum Fourier Transform (QFT)
• Variant of the Discrete Fourier Transform (DFT) that can be implemented on a quantum computer
• At the heart of Factoring and Order-Finding problems• QFT transforms state amplitudes to state amplitudes
– NOT qubits to qubits
Oct 25, 2005
17
QFT in Order Finding
x
y
H HR
Superpose QFTMeasure
x(y)
•States are measured according to their probability
•Many states at P produce the same x(y)
•QFT produces their frequency
•Probably answer reflects large number of states at P
PAnswer
Oct 25, 2005
18
Status of Computing
• Proof of concept factoring (2001)Chuang et al. 4-bit Shor algorithm implementation (2001)
•Ongoing ion-trap implementation effort
•Some optical lattice efforts
•Solid-state spins moving slowly
Oct 25, 2005
19
Ion Trap Investigations
• Done (per ARDA Roadmap April 2, 2004)– 2-qubit operations demonstrated– Long decoherence times in progress– 3-10 qubit operations started
• Proposed (individual researchers)– 10-20 qubit registers– Architectures with 1000 circulating qubits
• Possibility– 20 logical qubits (2-level error correct 1000 qubits)– 10 bit factoring
Oct 25, 2005
20
Optical Lattices
• Done (individual researchers)– 110 site lattice loaded from BEC with 200
atoms/site
• Proposed (individual researchers)– 8000 sites with CO2 lasers proposed by
Berkeley QuIST project• Filling factor 1/2
– Permits 80 logical qubits– Permits 40 bit factoring
Oct 25, 2005
21
Architectural Roadblocks
• Classical control– Large feature sizes for control lines mean large computers
• Wiring and corners– Moving qubits leads to decoherence
• Error correction– More check bits than data bits
• Cumulative effect– May need 100,000 times longer decoherence times than
required by operations alone (Balensiefer et. al, ISCA32, 2005, pp186-196).
Oct 25, 2005
22
Quantum Security
Oct 25, 2005
23
Quantum Key Distribution
• Use unforgability to detect eavesdropping
• Shared generation of secure key
• Extensive classical processing
Oct 25, 2005
24
Oct 25, 2005
25
Status of Key Distribution
QKDEndpoint
Encrypted Trafficvia Internet
QKDEndpoint
QKD Switch QKD Switch
QKD Switch
QKD Switch
QKD Repeater
PrivateEnclave
PrivateEnclave
Ultra-Long-Distance Fiber
End-to-End Key Distribution
BBN-AFRL-QuIST Network Rollout June 1, 2004
NEC 2-week demonstration May 31, 2005 (AFRL-QuIST inside?)
13kb/sec sifted key over 16km commercial access optical network
ID Quantique Turnkey System
Available throughout Switzerland June 05
Oct 25, 2005
26
Support Agencies
Oct 25, 2005
27
DARPA Mission
Focused strategic thrusts
Specific programs
Emphasize transition
Source: Bridging the Gap February 2005
Oct 25, 2005
28
DARPA Organization
Source: Bridging the Gap February 2005
Oct 25, 2005
29
QIP in DARPA Organization
Source: Bridging the Gap February 2005
Oct 25, 2005
30
NSF Mission
National Science Foundation Act of 1950 (Public Law 810507):
• To promote the progress of science; • to advance the national health, prosperity,
and welfare; • to secure the national defense; • and for other purposes.
Oct 25, 2005
31
NSF Organization
D ire c to ra te fo r B io lo g ica lS c ie n ces
D ire c to ra te fo r M a th e m a tica l& P h ysica l S cie n ces
D ire cto ra te fo r C o m pu te r &In fo rm atio n S cie n ce an d E n g in e ering
D ire c to ra te fo r S o cia l, B eh v io ra l& E con o m ic S c ie n ces
D ire cto ra te fo r E d u ca tion& H um a n R e sou rces
D ire c to ra te fo r G e o sc ie n ces
D ire cto ra te fo r E n g ine e ring O ffice o f P o la r P ro gra m s
O ffice o f In te gra tive A c tiv it ies
O ffice o f the D ire c to r
N a tion a l S cie n ceB o a rd
Administrative Offices
Oct 25, 2005
32
QIP in NSF Organization
D ire c to ra te fo r B io lo g ica lS c ie n ces
D ire c to ra te fo r M a th e m a tica l& P h ysica l S cie n ces
D ire cto ra te fo r C o m pu te r &In fo rm atio n S cie n ce an d E n g in e ering
D ire c to ra te fo r S o cia l, B eh v io ra l& E con o m ic S c ie n ces
D ire cto ra te fo r E d u ca tion& H um a n R e sou rces
D ire c to ra te fo r G e o sc ie n ces
D ire cto ra te fo r E n g ine e ring O ffice o f P o la r P ro gra m s
O ffice o f In te gra tive A c tiv it ies
O ffice o f the D ire c to r
N a tion a l S cie n ceB o a rd
Administrative Offices
Oct 25, 2005
33
CISE Mission
• to enable the United States to remain competitive in computing, communications, and information science and engineering;
• to promote understanding of the principles and uses of advanced computing, communications, and information systems in service to society; and
• to contribute to universal, transparent, and affordable participation in an information-based society.
CISE has three goals:
Oct 25, 2005
34
Desired Project Characteristics
• DARPA – Fast Results– Military need
– Technical challenges and plan for meeting them
– Transition plan
• NSF – Sustained Effort– Asks fundamental questions
– Maintains U.S. competitiveness
– Societal need
– Broad participation
Oct 25, 2005
35
DARPA Program Highlights
QKDEndpoint
Encrypted Trafficvia Internet
QKDEndpoint
QKD Switch QKD Switch
QKD Switch
QKD Switch
QKD Repeater
PrivateEnclave
PrivateEnclave
Ultra-Long-Distance Fiber
End-to-End Key Distribution
QuIST Network Rollout June 1, 2004
Chuang et al. 4-bit Shor algorithm implementation (2001)
Oct 25, 2005
36
NSF Program HighlightsInstitute for Quantum
Information at Caltech
CAREER Award: Yaoyun Shi at U. Michigan
Quantum Complexity and Polynomial Approximations of Boolean Functions
– Quantum and classical tradeoffs
– Classical simulation of quantum communication.
– Phase transition in biological signaling systems
– Education plan: theory of computation courses
Oct 25, 2005
37
Where Next for Communication?
• DARPA– Long range demonstrations between metronets– GtoA and GtoS demonstrations– ConOps for QKD
• NSF– New protocols– Security bounds
Oct 25, 2005
38
Where Next for Computation?
• New algorithms– Exponential speedups, please!– (Hashing outdoes unstructured search)
• New applications of existing algorithms– Pell’s equation– Random walk
• Scalable architectures– Controllable– Fault tolerant
• Medium-scale implementations– 10’s of qubits– Probably beyond NSF resources
Oct 25, 2005
39
The Near Future
• NSF budget is down 3% in 2005, looks flat
• DoD will need transitions beyond crypto
• New algorithms and protocols are needed for the next push– Scalable architectures too
Oct 25, 2005
40
The Want Ads
Oct 25, 2005
41
Program Directors Sought
• Numeric, Symbolic, Geometric computing
• Emerging Models and Technologies
• Interdisciplinary capability– Across cluster, division, NSF, and globally
Oct 25, 2005
42
Contact
• Vacancy announcements appear on www.nsf.gov• Meanwhile contact
Michael Foster
Division Director
Computing & Communication Foundations
National Science Foundation
4201 Wilson Boulevard
Arlington, VA 22230
703-292-8910