Clifford group from scratch Maris Ozols University of Waterloo - - PowerPoint PPT Presentation

clifford group from scratch
SMART_READER_LITE
LIVE PREVIEW

Clifford group from scratch Maris Ozols University of Waterloo - - PowerPoint PPT Presentation

Introduction Definition Single qubit case Order Generators Applications References Clifford group from scratch Maris Ozols University of Waterloo July 28, 2008 Introduction Definition Single qubit case Order Generators Applications


slide-1
SLIDE 1

Introduction Definition Single qubit case Order Generators Applications References

Clifford group from scratch

Maris Ozols

University of Waterloo

July 28, 2008

slide-2
SLIDE 2

Introduction Definition Single qubit case Order Generators Applications References

Outline

1 Introduction 2 Definition of the Clifford group Cn on n qubits 3 Clifford group C1 of a single qubit 4 Number of elements in Cn 5 Generators of Cn 6 Applications

slide-3
SLIDE 3

Introduction Definition Single qubit case Order Generators Applications References

Motivation

Everybody knows what the Clifford group is

slide-4
SLIDE 4

Introduction Definition Single qubit case Order Generators Applications References

Motivation

Everybody knows what the Clifford group is,

  • nly Maris doesn’t know. . .
slide-5
SLIDE 5

Introduction Definition Single qubit case Order Generators Applications References

Motivation

Everybody knows what the Clifford group is,

  • nly Maris doesn’t know. . .

I’m obsessed with symmetric structures in the Hilbert space

slide-6
SLIDE 6

Introduction Definition Single qubit case Order Generators Applications References

Motivation

Everybody knows what the Clifford group is,

  • nly Maris doesn’t know. . .

I’m obsessed with symmetric structures in the Hilbert space Clifford group has lots of applications

slide-7
SLIDE 7

Introduction Definition Single qubit case Order Generators Applications References

Motivation

Everybody knows what the Clifford group is,

  • nly Maris doesn’t know. . .

I’m obsessed with symmetric structures in the Hilbert space Clifford group has lots of applications I know the results, but I haven’t seen the proofs

slide-8
SLIDE 8

Introduction Definition Single qubit case Order Generators Applications References

Motivation

Everybody knows what the Clifford group is,

  • nly Maris doesn’t know. . .

I’m obsessed with symmetric structures in the Hilbert space Clifford group has lots of applications I know the results, but I haven’t seen the proofs Some folklore results with no proofs available

slide-9
SLIDE 9

Introduction Definition Single qubit case Order Generators Applications References

Pauli matrices

Single qubit

The set of Pauli matrices is P = {I, X, Y, Z}, where I = 1 0

0 1

  • ,

X = 0 1

1 0

  • ,

Y = 0 −i

i 0

  • ,

Z = 1 0

0 −1

  • .
slide-10
SLIDE 10

Introduction Definition Single qubit case Order Generators Applications References

Pauli matrices

Single qubit

The set of Pauli matrices is P = {I, X, Y, Z}, where I = 1 0

0 1

  • ,

X = 0 1

1 0

  • ,

Y = 0 −i

i 0

  • ,

Z = 1 0

0 −1

  • .

For n qubits

Pn = {σ1 ⊗ σ2 ⊗ · · · ⊗ σn | σi ∈ P} .

slide-11
SLIDE 11

Introduction Definition Single qubit case Order Generators Applications References

Pauli matrices

Single qubit

The set of Pauli matrices is P = {I, X, Y, Z}, where I = 1 0

0 1

  • ,

X = 0 1

1 0

  • ,

Y = 0 −i

i 0

  • ,

Z = 1 0

0 −1

  • .

For n qubits

Pn = {σ1 ⊗ σ2 ⊗ · · · ⊗ σn | σi ∈ P} .

Vector space structure

The group Pn/U(1) is isomorphic to a vector space over F2 with dimension 2n via identification Z Y | | I X multiply ⇐ ⇒ (0, 1) (1, 1) | | (0, 0) (1, 0) add

slide-12
SLIDE 12

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

slide-13
SLIDE 13

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1.

slide-14
SLIDE 14

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

slide-15
SLIDE 15

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in P ∗

n have eigenvalues ±1 with equal multiplicity.

slide-16
SLIDE 16

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

slide-17
SLIDE 17

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

slide-18
SLIDE 18

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X,

slide-19
SLIDE 19

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X,

slide-20
SLIDE 20

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X,

slide-21
SLIDE 21

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

slide-22
SLIDE 22

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

You cannot

slide-23
SLIDE 23

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

You cannot

X → I,

slide-24
SLIDE 24

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

You cannot

X → I, X → iX.

slide-25
SLIDE 25

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take Paulis to Paulis via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

You cannot

X → I, X → iX.

slide-26
SLIDE 26

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take ±P ∗

n to ±P ∗ n via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

You cannot

X → I, X → iX.

slide-27
SLIDE 27

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition (sloppy)

Unitaries that take ±P ∗

n to ±P ∗ n via conjugation.

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

You cannot

X → I, X → iX.

Global phase

U and eiϕU act identically, i.e., UMU† = (eiϕU)M(eiϕU)†.

slide-28
SLIDE 28

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Definition

The Clifford group Cn on n qubits is Cn =

  • U ∈ U(2n) | σ ∈ ±P ∗

n ⇒ UσU† ∈ ±P ∗ n

  • /U(1).

Eigenvalues

The eigenvalues of X, Y , Z are ±1. Let P ∗

n = Pn \ {I⊗n}.

All matrices in ± P ∗

n have eigenvalues ±1 with equal multiplicity.

You can

X → −X, e.g., ZXZ = −X, X ⊗ I → X ⊗ X, e.g., CNOT(X ⊗ I)CNOT † = X ⊗ X.

You cannot

X → I, X → iX.

Global phase

U and eiϕU act identically, i.e., UMU† = (eiϕU)M(eiϕU)†.

slide-29
SLIDE 29

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

slide-30
SLIDE 30

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

Restrictions

Conjugation must preserve the structure of Pauli matrices.

slide-31
SLIDE 31

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

Restrictions

Conjugation must preserve the structure of Pauli matrices. Y = iXZ, thus UY U† = i(UXU†)(UZU†),

slide-32
SLIDE 32

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

Restrictions

Conjugation must preserve the structure of Pauli matrices. Y = iXZ, thus UY U† = i(UXU†)(UZU†), U(−X)U† = −UXU † and similarly for Z.

slide-33
SLIDE 33

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

Restrictions

Conjugation must preserve the structure of Pauli matrices. Y = iXZ, thus UY U† = i(UXU†)(UZU†), U(−X)U† = −UXU † and similarly for Z. Thus it is enough to specify where X and Z go.

slide-34
SLIDE 34

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

Restrictions

Conjugation must preserve the structure of Pauli matrices. Y = iXZ, thus UY U† = i(UXU†)(UZU†), U(−X)U† = −UXU † and similarly for Z. Thus it is enough to specify where X and Z go. However, since X and Z anti-commute, so must UXU† and UZU†.

slide-35
SLIDE 35

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

Restrictions

Conjugation must preserve the structure of Pauli matrices. Y = iXZ, thus UY U† = i(UXU†)(UZU†), U(−X)U† = −UXU † and similarly for Z. Thus it is enough to specify where X and Z go. However, since X and Z anti-commute, so must UXU† and UZU†.

All possibilities

X can go to any element of ±P ∗

1 ,

Z can go to any element of ±P ∗

1 \

  • ±UXU†

.

slide-36
SLIDE 36

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Single qubit

±P ∗

1 = {±X, ±Y, ±Z}.

Restrictions

Conjugation must preserve the structure of Pauli matrices. Y = iXZ, thus UY U† = i(UXU†)(UZU†), U(−X)U† = −UXU † and similarly for Z. Thus it is enough to specify where X and Z go. However, since X and Z anti-commute, so must UXU† and UZU†.

All possibilities

X can go to any element of ±P ∗

1 ,

Z can go to any element of ±P ∗

1 \

  • ±UXU†

.

Group order

|C1| = 6 · 4 = 24.

slide-37
SLIDE 37

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Clifford group rotations

slide-38
SLIDE 38

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Clifford group rotations

slide-39
SLIDE 39

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Clifford group rotations

slide-40
SLIDE 40

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Clifford group rotations

slide-41
SLIDE 41

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Cuboctahedron

slide-42
SLIDE 42

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Cuboctahedron

slide-43
SLIDE 43

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Cuboctahedron Poll

Guess what’s the value of |C2|?

slide-44
SLIDE 44

Introduction Definition Single qubit case Order Generators Applications References

Clifford group C1

Cuboctahedron Poll

Guess what’s the value of |C2|? Answer: |C2| = 11520.

slide-45
SLIDE 45

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

It is enough to specify where Xi and Zi go for all i ∈ {1, . . . , n}.

slide-46
SLIDE 46

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

It is enough to specify where Xi and Zi go for all i ∈ {1, . . . , n}. All X’s and Z’s commute, except Xi and Zi that anti-commute: X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

slide-47
SLIDE 47

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

It is enough to specify where Xi and Zi go for all i ∈ {1, . . . , n}. All X’s and Z’s commute, except Xi and Zi that anti-commute: X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Claim

Each matrix in ±P ∗

n commutes (anti-commutes) with exactly half

  • f Pauli matrices Pn.

Proof.

Let σ ∈ ±P ∗

n and k be a position where σ does not contain I. All

Paulis that anti-commute with σ can be constructed as follows:

slide-48
SLIDE 48

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

It is enough to specify where Xi and Zi go for all i ∈ {1, . . . , n}. All X’s and Z’s commute, except Xi and Zi that anti-commute: X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Claim

Each matrix in ±P ∗

n commutes (anti-commutes) with exactly half

  • f Pauli matrices Pn.

Proof.

Let σ ∈ ±P ∗

n and k be a position where σ does not contain I. All

Paulis that anti-commute with σ can be constructed as follows: put any of I, X, Y , Z at each position other than k,

slide-49
SLIDE 49

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

It is enough to specify where Xi and Zi go for all i ∈ {1, . . . , n}. All X’s and Z’s commute, except Xi and Zi that anti-commute: X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Claim

Each matrix in ±P ∗

n commutes (anti-commutes) with exactly half

  • f Pauli matrices Pn.

Proof.

Let σ ∈ ±P ∗

n and k be a position where σ does not contain I. All

Paulis that anti-commute with σ can be constructed as follows: put any of I, X, Y , Z at each position other than k, fill the kth position in any of two possible ways so that the

  • btained matrix anti-commutes with σ.
slide-50
SLIDE 50

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

slide-51
SLIDE 51

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Counting

Where can U ∈ Cn send Xn and Zn?

slide-52
SLIDE 52

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Counting

Where can U ∈ Cn send Xn and Zn? Xn can go to any element of ±P ∗

n, i.e., 2(4n − 1) choices,

slide-53
SLIDE 53

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Counting

Where can U ∈ Cn send Xn and Zn? Xn can go to any element of ±P ∗

n, i.e., 2(4n − 1) choices,

Zn can go to any element of ±P ∗

n that anti-commutes with

UXnU†, i.e., 2|Pn|

2

= 4n choices.

slide-54
SLIDE 54

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Counting

Where can U ∈ Cn send Xn and Zn? Xn can go to any element of ±P ∗

n, i.e., 2(4n − 1) choices,

Zn can go to any element of ±P ∗

n that anti-commutes with

UXnU†, i.e., 2|Pn|

2

= 4n choices. Similarly for the next pair (Xn−1, Zn−1), just replace n by n − 1.

slide-55
SLIDE 55

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Restrictions

X1 X2 . . . Xn−1 Xn | | | | Z1 Z2 . . . Zn−1 Zn

Counting

Where can U ∈ Cn send Xn and Zn? Xn can go to any element of ±P ∗

n, i.e., 2(4n − 1) choices,

Zn can go to any element of ±P ∗

n that anti-commutes with

UXnU†, i.e., 2|Pn|

2

= 4n choices. Similarly for the next pair (Xn−1, Zn−1), just replace n by n − 1.

Result

|Cn| =

n

  • j=1

2(4j − 1) · 4j = 2n2+2n

n

  • j=1

(4j − 1).

slide-56
SLIDE 56

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

How does it grow?

n |Cn| 1 24 2 11520 3 92897280 4 12128668876800 5 25410822678459187200

slide-57
SLIDE 57

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

How does it grow?

n |Cn| 1 24 2 11520 3 92897280 4 12128668876800 5 25410822678459187200 This is 1

8 times “Sloane’s A003956”.

slide-58
SLIDE 58

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

How does it grow?

n |Cn| 1 24 2 11520 3 92897280 4 12128668876800 5 25410822678459187200 This is 1

8 times “Sloane’s A003956”.

Upper bound

|Cn| ≤ 2n2+2n

n

  • j=1

4j = 22n2+3n.

slide-59
SLIDE 59

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

slide-60
SLIDE 60

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Their definition

Calderbank R.A., Rains E.M., Shor P.W., Sloane N.J.A., Quantum Error Correction Via Codes Over GF(4), arXiv:quant-ph/9608006v5.

slide-61
SLIDE 61

Introduction Definition Single qubit case Order Generators Applications References

Order of Cn

Their definition

Calderbank R.A., Rains E.M., Shor P.W., Sloane N.J.A., Quantum Error Correction Via Codes Over GF(4), arXiv:quant-ph/9608006v5.

Explanation of factor 8

They assume that H, P ∈ Cn, i.e., they define Cn as the group generated by H, P, and CNOT. Thus they get 8 times more, since ηI ∈ Cn, where η = 1+i

√ 2 is the 8th root of unity.

slide-62
SLIDE 62

Introduction Definition Single qubit case Order Generators Applications References

Generators of Cn

Theorem

The Clifford group Cn is generated by H, P, and CNOT: H =

1 √ 2

1 1

1 −1

  • ,

P = 1 0

0 i

  • ,

CNOT = 1 0 0 0

0 1 0 0 0 0 0 1 0 0 1 0

  • .
slide-63
SLIDE 63

Introduction Definition Single qubit case Order Generators Applications References

Generators of Cn

Theorem

The Clifford group Cn is generated by H, P, and CNOT: H =

1 √ 2

1 1

1 −1

  • ,

P = 1 0

0 i

  • ,

CNOT = 1 0 0 0

0 1 0 0 0 0 0 1 0 0 1 0

  • .

More precisely, Cn = Hi, Pi, CNOTij /U(1).

slide-64
SLIDE 64

Introduction Definition Single qubit case Order Generators Applications References

Generators of Cn

Theorem

The Clifford group Cn is generated by H, P, and CNOT: H =

1 √ 2

1 1

1 −1

  • ,

P = 1 0

0 i

  • ,

CNOT = 1 0 0 0

0 1 0 0 0 0 0 1 0 0 1 0

  • .

More precisely, Cn = Hi, Pi, CNOTij /U(1).

Proof.

It is easy to verify that C1 = H, P /U(1). Use induction on n.

slide-65
SLIDE 65

Introduction Definition Single qubit case Order Generators Applications References

Generators of Cn

Proof (continued).

Let U ∈ Cn+1. Since X1 and Z1 anti-commute, so do UX1U† and UZ1U†. We can permute qubits and apply elements of C1 so that UX1U† = X ⊗ M′, UZ1U† = Z ⊗ N′. for some M′, N′ ∈ ±Pn.

slide-66
SLIDE 66

Introduction Definition Single qubit case Order Generators Applications References

Generators of Cn

Proof (continued).

Let U ∈ Cn+1. Since X1 and Z1 anti-commute, so do UX1U† and UZ1U†. We can permute qubits and apply elements of C1 so that UX1U† = X ⊗ M′, UZ1U† = Z ⊗ N′. for some M′, N′ ∈ ±Pn. Let U(|0 ⊗ |ψ) =

1 √ 2(|0 ⊗ |ψ0 + |1 ⊗ |ψ1).

Define U′ by U′ |ψ = |ψ0. One can show that U′ ∈ Cn.

slide-67
SLIDE 67

Introduction Definition Single qubit case Order Generators Applications References

Generators of Cn

Proof (continued).

Let U ∈ Cn+1. Since X1 and Z1 anti-commute, so do UX1U† and UZ1U†. We can permute qubits and apply elements of C1 so that UX1U† = X ⊗ M′, UZ1U† = Z ⊗ N′. for some M′, N′ ∈ ±Pn. Let U(|0 ⊗ |ψ) =

1 √ 2(|0 ⊗ |ψ0 + |1 ⊗ |ψ1).

Define U′ by U′ |ψ = |ψ0. One can show that U′ ∈ Cn. Then we can implement U as follows:

  • H
  • /

U′ / N′ / M′ /

slide-68
SLIDE 68

Introduction Definition Single qubit case Order Generators Applications References

Stabilizer formalism

slide-69
SLIDE 69

Introduction Definition Single qubit case Order Generators Applications References

Stabilizer formalism

Who doesn’t know that the stabilizer formalism is?

slide-70
SLIDE 70

Introduction Definition Single qubit case Order Generators Applications References

Gottesman-Knill theorem

Schr¨

  • dinger vs. Heisenberg
slide-71
SLIDE 71

Introduction Definition Single qubit case Order Generators Applications References

Gottesman-Knill theorem

Schr¨

  • dinger vs. Heisenberg

Schr¨

  • dinger picture: quantum states evolve in time,
slide-72
SLIDE 72

Introduction Definition Single qubit case Order Generators Applications References

Gottesman-Knill theorem

Schr¨

  • dinger vs. Heisenberg

Schr¨

  • dinger picture: quantum states evolve in time,

Heisenberg picture: operators evolve in time.

slide-73
SLIDE 73

Introduction Definition Single qubit case Order Generators Applications References

Gottesman-Knill theorem

Schr¨

  • dinger vs. Heisenberg

Schr¨

  • dinger picture: quantum states evolve in time,

Heisenberg picture: operators evolve in time.

Theorem (Gottesman-Knill)

Any quantum computation involving only: measurements in standard basis, Clifford group gates (conditioned on classical bits, e.g., measurement outcomes) can be perfectly simulated in polynomial time on a probabilistic classical computer.

slide-74
SLIDE 74

Introduction Definition Single qubit case Order Generators Applications References

Gottesman-Knill theorem

Schr¨

  • dinger vs. Heisenberg

Schr¨

  • dinger picture: quantum states evolve in time,

Heisenberg picture: operators evolve in time.

Theorem (Gottesman-Knill)

Any quantum computation involving only: measurements in standard basis, Clifford group gates (conditioned on classical bits, e.g., measurement outcomes) can be perfectly simulated in polynomial time on a probabilistic classical computer.

CHP (CNOT-Hadamard-Phase)

Program in C written by Aaronson and Gottesman to simulate such circuits. Can easily handle up to 3000 qubits!

slide-75
SLIDE 75

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

slide-76
SLIDE 76

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

slide-77
SLIDE 77

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

Proof.

. . .

slide-78
SLIDE 78

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

Proof.

. . . RCM-lattices . . .

slide-79
SLIDE 79

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

Proof.

. . . RCM-lattices . . . natural module . . .

slide-80
SLIDE 80

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

Proof.

. . . RCM-lattices . . . natural module . . . inertia group . . .

slide-81
SLIDE 81

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

Proof.

. . . RCM-lattices . . . natural module . . . inertia group . . . is ramified . . .

slide-82
SLIDE 82

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

Proof.

. . . RCM-lattices . . . natural module . . . inertia group . . . is ramified . . . which is a contradiction.

slide-83
SLIDE 83

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Mathematicians have shown that. . .

Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2.

In other words

Let m ≥ 1. Then Cm together with any other gate not in Cm form a universal set of quantum gates.

Proof.

. . . RCM-lattices . . . natural module . . . inertia group . . . is ramified . . . which is a contradiction.

Question

Is there an elementary proof for this?

slide-84
SLIDE 84

Introduction Definition Single qubit case Order Generators Applications References

Universal set of quantum gates

Another zero-knowledge proof

slide-85
SLIDE 85

Introduction Definition Single qubit case Order Generators Applications References

References

slide-86
SLIDE 86

Introduction Definition Single qubit case Order Generators Applications References

Clifford group

Calderbank R.A., Rains E.M., Shor P.W., Sloane N.J.A., Quantum Error Correction Via Codes Over GF(4), arXiv:quant-ph/9608006v5. Nebe G., Rains E.M., Sloane N.J.A., The Invariants of the Clifford Groups, arXiv:math/0001038v2. Planat M., Jorrand P., Group theory for quantum gates and quantum coherence, arXiv:0803.1911v2 [quant-ph].

slide-87
SLIDE 87

Introduction Definition Single qubit case Order Generators Applications References

H, P, and CNOT generate Cn

Gottesman D., Stabilizer Codes and Quantum Error Correction, PhD thesis, arXiv:quant-ph/9705052v1. Gottesman D., A Theory of Fault-Tolerant Quantum Computation, arXiv:quant-ph/9702029v2. Nielsen M.A., Chuang I.L., Quantum Computation and Quantum Information, Cambridge University Press, 2000.

slide-88
SLIDE 88

Introduction Definition Single qubit case Order Generators Applications References

Gottesman-Knill theorem

Gottesman D., The Heisenberg Representation of Quantum Computers, arXiv:quant-ph/9807006v1. Aaronson S., Gottesman D., Improved Simulation of Stabilizer Circuits, arXiv:quant-ph/0406196v5.

slide-89
SLIDE 89

Introduction Definition Single qubit case Order Generators Applications References

Thank you for your attention!