. . ..

Wanted: Schodingers Cat: Dead and Alive

. .

March 7, 2014 Wichita State University

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

1 / 31

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

2 / 31

Quantum Computation

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

3 / 31

Nondeterministic classical states work like this:

Someone puts either a black marble in the box with probability p, or a white marble in the box with probability 1 p. The state of the box with regard to the color of the marble it contains is either black with probability p, or white probability (1 p). You look in the box to measure the state of the box in regard to the color of the marble. Suppose you see a black marble. Then there was a marble there along and it was black.

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

4 / 31

Quantum States:
A simple example: One bit. We physically represent bits many ways. One way: circularly polarized photons. 3D glasses at the movies.

Clockwize polarization can be regarded as representing bit value 0 and counterclockwize polarization regarded as representing 1.
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

5 / 31

Spin:

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

6 / 31

Quantum States:

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

7 / 31

Superposition:
The circular polarization property of a photon is one of implementing a qubit : its state, i.e. its polarization direction, is a quantum state: a 0 + b 1, called a wave function. a, b are square roots of probabilities with phase - called amplitudes. Wave functions can with combine constructively and destructively. a 0 + b 1 = ei p 0 + ei (1 p) 1 The squares of the absolute values of a and b are probabilities. The absolute value operation removes phase. e.g. ei p = p
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

8 / 31

The rules (of projective measurement)

Observe a qubit thats in state a 0 + b 1. The result of your observation will be either 0 or 1. The probability of getting 0 is |a|2 and of getting 1 is |b|2 . (Wave function collapse.) After the observation the photon is either in state 0 or in state 1 depending on whether 0 or 1 was observed. Important Question: How do we know the state really wasnt 0 or 1, like the marble in the box, before we observed it?

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

9 / 31

2 photons: Entanglement
a 00 + b 01 + c 10 + d 11 Suppose we have two photons, the rst in state a 0 + b 1 and the second in state c 0 + d 1. The combined system is in the separable state (a 0 + b 1) (c 0 + d 1) = ac 00 + ad 01 + bc 10 + bd 11 What about states like 1 1 00 + 11 2 2 The quantum states of the separate qubits arent there. Therefore those qubits do not separately exist? (Quantum Leibniz)
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

10 / 31

Quantum Teleportation
The Setup 1: On an apparatus in her lab in New York, Alice has a qubit in some quantum state a 0 + b1. She does not know the quantum state of her qubit. She wants to send the qubit to Bob who is in Paris. Specically, she wants to cause the quantum state of the qubit storage apparatus in Bobs lab in Paris to become a 0 + b 1. The Setup 2: Alice and Bob have prepared a pair of entangled qubits in (Bell or EPR ) state 1 1 00 + 11 2 2 on a pair of single qubit storage devices, one of which they each took to their labs. We have suggested there are no separate qubit states stored in their qubits from the entangled pair.
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

11 / 31

Quantum Teleportation
Initially, there are three qubit storage devices in play: (1) holds Alices qubit that she wants to teleport; (2) Alices qubit participating in the entangled pair; (3) Bobs qubit participating in the entangled pair. AToBeSent AEntangled BEntangled The state of the three devices is partially entangled like this: 1 1 (a0 + b1) ( 00 + 11) 2 2 = a a b b 000 + 011 + 100 + 111 2 2 2 2
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

12 / 31

Quantum Teleportation
Alice pretends that the state of her qubit part of the entangled pair really exists and sends her qubit state that she wants to teleport and her qubit state from the entangled pair through a controlled-not quantum gate. The state of the three devices becomes: a b b a 000 + 011 + 110 + 101 2 2 2 2 Classical controlled-not gate:
Control

0 0 1
. . .

0 0 1
. . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

13 / 31

Quantum controlled-not gate

Qubit values 0 and 1, respectively: [ ] [ ] 1 0 0 1 The quantum controlled-not gate must map pairs of inputs to pairs of outputs in the same that the classical controlled-not gate does. Pairs of quantum states on separate, distinguishable components result in tensor products of the separate states. In matrix form we take Kronecker products.

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

14 / 31

Distinguishable multi-partite quantum systems

Suppose we have two quantum bits a 0 + b1 In matrix notation [ a0 + b1 = a [ c0+d1=c 1 0 1 0 ] +b ] +d [ [ 0 1 0 1 ] = ] = [ [ a b c d ] ] c 0 + d1

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

15 / 31

Distinguishable multi-partite quantum systems

The combined system of two qubits is in state ac ad (a 0 + b 1) (c 0 + d 1) = bc bd

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

16 / 31

Distinguishable multi-partite quantum systems

0 0 0 1 0 0 1 0

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

17 / 31

Distinguishable multi-partite quantum systems

Classical (sequential) computation: A sequence of small changes in the computational state; i.e. a state is given by the values of all of the variables involved. State change consists of changing the value of one of the variables. Quantum computation: State is multipartite: some components of the state a quantum states, others classical states. State change is consists of classical changes to classical components, or unitary changes to quantum components, or measurements of quantum components.
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

18 / 31

Quantum Teleportation
We left Alices and Bobs three qubit multipartite system in state a a b b 000 + 011 + 110 + 101 2 2 2 2 Alice then sends her qubit (the one to be teleported) through a Hadamard gate: 1 1 0 ( 0 + 1) 2 2 1 1 1 ( 0 1) 2 2

Substituting on the rst qubit, the state of the three partially entangled devices is now: 1 1 a 1 1 a ( 0 + 1)00 + ( 0 + 1)11 2 2 2 2 2 2 b 1 1 b 1 1 + ( 0 1)10 + ( 0 1)01 2 2 2 2 2 2
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

19 / 31

Quantum Teleportation

which simplies to a a a a 000 + 100 + 011 + 111 2 2 2 2 b b b b + 010 110 + 001 101 2 2 2 2 (and now for more quantum wierdness): = 1 1 1 1 00(a 0 + b 1)+ 01(a 1 + b 0)+ 10(a 0 b 1)+ 11(a 1 b 0) 2 2 2 2

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

20 / 31

Quantum Teleportation
Alice performs a projective measurement on both of her qubits which will cause a wave function partial collapse of the entire three-qubit system shared by Alice and Bob. Suppose for example that Alices qubit to sent is measured to be 1 and the qubit originally entangled with Bobs is measured to be 0. Then the quantum state of the system collapses to 10(a 0 b 1) Bobs qubit is now in state a 0 b 1. He doesnt know a and b. Alice emails Bob to tell him the result of her measurement. Bob then knows that if he applies a phase-ip operation (i.e. 0 0 and 1 1) he will have Alices qubit (value) in the device in his lab a0 + b1
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

21 / 31

Factoring pure quantum states

where

xFn 2

|ax |2 = 1

Consider the ax arranged in a 2 2 . . . 2 array, i.e. an n dimensional hypermatrix T . Then is factorable as a tensor product of individual qubit states if, and only if, the determinant of every 2 2 submatrix of T vanishes.
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

22 / 31

Factoring pure quantum states

Entangled?

a a a a 000 + 100 + 011 + 111 2 2 2 2 b b b b + 010 110 + 001 101 2 2 2 2

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

23 / 31

Factoring pure quantum states

b/2 010

-b/2 110

b/2 001

-b/2 101

a/2 000
Howard A. Blair (Syracuse University) QuantumComputing

a/2 100
. . . . . .

03/07/14

24 / 31

SQIL

qubit p,q,r; // All qubits are assumed to be randomly // initialized in some quantum state. bit u,v,w; //u,v,w are initialized to 0. measure q in w; // After execution the state of q and v // is either 0 and 0, or 1 and 1. if (w = 1) // Notice that the state transition performed by // this branching command is not unitary. then NOT(q); // The state of q after this branching command is // is 0. NOT(x0 + y1) = y0 + x1

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

25 / 31

SQIL
measure r in w; if (w = 1) then NOT(r); // At this point the state of the qubit pair (q,r) is // 00. The combined state of all three qubits is // a 000 + b 100 BELL(q,r); // BELL(x,y) can be any 2-qubit unitary that maps // 00 to the Bell state // (1/sqrt2) 00 + (1/sqrt2) 11. // To see the combined state of all three qubits, // multiply (a 0 + b textbf1)((1/sqrt2) 00 + (1/sqrt2) 11) CNOT(p,q); // CNOT(00) = 00. CNOT(01) = 01. // CNOT(10) = 11. CNOT(11) = 10.
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

26 / 31

SQIL
Hadamard(p); measure p in u; measure q in v; // Four possibilities for u, v if (u = 0) and (v = 1) then NOT(r) else if (u = 1) and (v = 0) then PhaseFlipPi(r) //PhaseFlipPi(x 0 + y 1) = x 0 - y 1 else { PhaseFlipPi(r); NOT(r) }
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

27 / 31

How do we know quantum states are not classical?

Answer: Kochen-Specker, Bell/CSHS inequality, ... Simon Kochen, Ernst Specker, 1968 J.F. Clauser, M.A. Horne, A. Shimony, R.A. Holt, 1969

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

28 / 31

Bell
Alice and Bob are given a large collection of entangled pairs of qubits, where each pair is entangled in a phase-ipped Bell state 1 1 01 10 2 2 For each pair, Alice has one of the qubits, Bob the other. Alice prepares two apparatuses to perform a projective measurement on each of her qubits that will return +1 corresponding to bit-value of 0 and -1 corresponding to a bit-value of 1. Specically, Alice prepares the observables [ ] [ ] 1 0 0 1 Z = X = 0 1 1 0
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

29 / 31

Bell
Bob prepares similar but more elaborate observables 1 S = (Z X ) 2 1 T = (Z X ) 2

Bob leaves for our Martian colony with his qubits and apparatuses. At a prearranged time when Mars and Earth are about 20 light-minutes apart, Alice and Bob each measure their qubits using their prepared observables. We assume their actions take much less than 20 minutes and are therefore relativistically causally separated. ? What is the expected value of the random variable V = ZA SB + XA SB + XA TB ZA TB
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

30 / 31

Bell
If we assume that Alices and Bobs qubits had denite probabilistically determined 0 or 1 valued states prior to measurement, then we obtain, independently of the probability distribution, the rst Bell inequality : E(V ) = E(ZA SB + XA SB + XA TB ZA TB ) 2 If we calculate the expected values of the quantum mechanical observables on the entangled qubit pairs, we obtain for each of the component observables 1 E(ZA SB ) = E(XA SB ) = E(XA TB ) = E(ZA TB ) = 2 Hence E(V ) = 2 2
. . . . . .

Howard A. Blair (Syracuse University)

QuantumComputing

03/07/14

31 / 31