Dung Nguyen Chicago 19 th January Content Motivation Quantum bit qubit vs Classical bit bit Quantum Computation Quantum Communication Conclusion Motivation The end of Moores law scaling in silicon because of quantum effects of particle at scale smaller than 7nm ID: 555667
Download Presentation The PPT/PDF document "Introduction to quantum computing and qu..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Introduction to quantum computing and quantum information
Dung Nguyen
Chicago 19
th
JanuarySlide2
Content
Motivation
Quantum bit (qubit) vs Classical bit (bit)
Quantum Computation
Quantum Communication
ConclusionSlide3
Motivation
The end of Moore’s law scaling in silicon (because of quantum effects of particle at scale smaller than 7nm).
There are only three more generations of Intel processor based on current silicon technology.
There could be new improvements for silicon technology (3D stacking), but the issues of cooling and energy consuming are hard to pass
The new carbon nanotubes computer is questionable (easy to have defects that can ruin an entire chip)Slide4
Motivation
The classical computer can not simulate
efficiently
quantum systems (an unproven conjecture).
Quantum computer can simulate quantum systems, so there is a possible solution for strong correlated
correlated problems (high Tc superconductor) using quantum computation. Quantum computer can solve the hard problems of classical computing using computational power inherent in quantum mechanics. Quantum algorithms can solve some of the NP problems (but not NP-complete) like searching.
Slide5
Motivation
The security of quantum communication (we will talk about it)
Quantum computation can break the classical cryptography (finding prime factors) using reasonable time and energy consumption.Slide6
Quantum bit (Qbit) vs Classical bit (bit)
Classical bit and classical computer
Classical bit take value 0 or 1
We can consider logic gates as the function
Example of
Slide7
Quantum bit (Qbit) vs Classical bit (bit)
Classical bit
Example of
Slide8
Quantum bit (Qbit) vs Classical bit (bit)
Universal gates
Any
can be built up using NAND gates and Fan-out.
Example:
Slide9
Quantum bit (Qubit) vs Classical bit (bit)
Stronger Universality statement (Strong Church-Turing thesis)
Any algorithmic process (in Turing’s definition)can be simulated efficiently using a probabilistic Turing machine (loosely speaking a computer which is built up by logic gates and can generate random numbers)
Every Turing machine has equal power (
polynomially
).
The idea behind is a Turing machine can simulate an other Turing machine within a polynomially bounded overhead in
time. Slide10
Quantum bit (Qubit) vs Classical bit (bit)
Quantum logic bit (qubit)
A single qubit can be described by a vector two dimensional complex Hilbert space
When we
mesuare
a qubits, we either get the result 0, with probability
,or the result 1, with the probability
. The normalization condition
, we can rewrite the
qbit
as
We can represent it by the Bloch sphere.
The single qubit can be represented by 2 real numbers
and
.
Slide11
Quantum bit (Qubit) vs Classical bit (bit)
Because of quantum
decoherence
,
the quantum logic bit usually is not a single physical qubit.
The single physical qubit can be constructed by single atom, by photon polarizations, by spin direction of single electron, superconducting circuit (based on Josephson effect)
The orthonormal basis can be chosen arbitrary. We can chose the basis
and
Slide12
Quantum bit (Qubit) vs Classical bit (bit)
How much information we can store in a single qubit if we measure it? There are an infinite number of points on the Bloch sphere, could we store infinite binary expansion of
or
?
NO! Because of behavior of qubit when observed, when we measure a qubit, it gives the value either 0 or 1. From a single measurement we only can have a single bit information.
If you want to have all information of
and
, you need to measure
infinite
identically qubits, which is impossible in general due to “no-cloning” theorem (we will talk about it latter).
Slide13
Quantum bit (Qubit) vs Classical bit (bit)
How much information can be stored in a qubit, if we don’t measure it? There is something conceptually important here, we can do the computation without doing measurement in the middle step, we do the measurement at the end.
A qubit contain hidden quantum information, and this hidden information grows exponentially with the number of qubit. Slide14
Quantum bit (Qubit) vs Classical bit (bit)
Multiple qubits
If we have 2 qubits, we have four computational basis states
The state vector that describing the 2 qubit system is
For a 2 qubit system, we can measure subset of the qubits, if we measure the first qubit, and get the value 0 with probability
, and the post measurement state (collapsed state)
Slide15
Quantum bit (Qubit) vs Classical bit (bit)
The 2 qubit system can be described with 4 complex number, however if we consider normalization and ignore arbitrary phase factor, we only need 6 real numbers.
More generally, we may consider a system of n qubits, we need a vector in
to represent a state (but in practice, we use
complex numbers ). For
n=300
, the dimension of our vector space is bigger than total number of atoms in the visible Universe. The hidden information is huge.
Could we use this hidden information to do computation (quantum computation)?
YES!
Slide16
Quantum computation
Single qubit gate
is any
unitary transformation act on the vector space that represent a qubit, it can be considered as an evolution matrix of a time dependent Hamiltonian
We do have access to hidden information, not just
and
states.
Examples:
Quantum NOT gate
Phase gate
Slide17
Quantum computation
Z gate
Hadamard
gate
Transformation of a qubit state under action of qubit logic gates
There are infinite many 2 by 2 unitary matrices, do we need infinitely many single qubit logic gate? NO!Slide18
Quantum computation
Universal single qubit gates
We know that any unitary transformation in 2 dimension can be decomposed in to Euler rotations
So with any unitary transformation, we can built it up from 4 gates? Not really, since
are real numbers, we still need infinite number of gate to built up exactly any unitary transformation.
However with in an approximation
|U|
w
here
is the transformation built from quantum gates.
We need only some certain special fixed value of
, and the price is we need more than 4 gates to built up
.
Slide19
Quantum computation
Example for an arbitrary real number
, and a fixed irrational number
, we always can find a integer number m such that
Thus and
gate can be built up from m
gates.
This is just an example to show that we can have a set of universal single qubit gates. In practice, we may use different set of gates as the universal single qubit gates.
Slide20
Quantum computation
Multiple qubit quantum gate
We have a special 2 qubit quantum gate is controlled-NOT or CNOT gate act on the basis of 2 qubit state
.
Or in matrix form
Where
is addition modulo 2.
Slide21
Quantum bit (Qubit) vs Classical bit (bit)
Universal multiple qubits gates
We can prove that any multiple qubits transformation can be constructed from CNOT gates and single qubit gates.
For an approximation
, any multiple qubits gate can be built from CNOT gate,
Hadamard
gate and
gate.
Thus we have
CNOT gate,
Hadamard
gate and
gate are the set of universal gates for multiple qubits.
We have quantum Church-Turing thesis: Every quantum Turing machine has equal power (polynomially).
Slide22
Quantum Computation
No-cloning theorem
We can not do quantum copy on a qubit
Proof:
If we can find an unitary operator C (clone operator) such that with any normalized qubit state
, we have
Thus we have
=
This can not be true for general state.
There is no quantum Fan-out.
Slide23
Quantum Computation
Control quantum gate:
Consider U is any unitary matrix acting on
qubits. Define a controlled-U gate as a extension of controlled-NOT gate with an extra control qubit. If the control qubit is set to 0, then nothing happens to the
target qubits. If the control qubit is set to 1 then the gate U is applied to the target qubits
Slide24
Quantum Computation
Can quantum computer do the classical computation?
YES!
We require the input and output are both classical signals (either
or
)
We can construct the classical gates using the quantum circuit (indeed we make the reversible version of the classical gate) named
Toffoli
gate
Slide25
Quantum Computation
We just need to show that we can construct reversible version of NAND and Fan-out
Generate random number is the property that is inherited from quantum mechanics. Slide26
Quantum computation
Can quantum computer do better?
YES!
One of the most important example is quantum parallelism
We consider a function on classical bit
We can construct quantum gate
gate with the specific input
(We can easily show that
is unitary by writing down the matrix form in the canonical basis)
Slide27
Quantum Computation
We see that the output state is
The different terms contain information about both
and
.
Generally, if we want to calculate the function of n bit
, we can make a quantum circuit including n+1 input and n+1 output qubits, such that the output state has the information of
with
take
value simultaneously
But what information we can get from that final state?
Slide28
Quantum Computation
Let investigate it more
Using this quantum circuit, at the end we found that the final state
When
mesure
the first qubits, we can determine
, quantum circuit gives us a
global property
of function
with only one evaluation. The power is exponentially faster (in the number of bit) in comparison to classical computer.
Slide29
Quantum Computation
Exploit the quantum parallelism, we can solve some of NP problems, we can finish the computation
polynomially
in time rather than exponentially in time.
Some of the most famous problem are quantum searching and prime factoring (Shor’s algorithm). The prime factoring is the key of classical cryptography
If we could do quantum computing now, the online trading is dead. Slide30
Quantum Communication
Using qubit, we can store and transfer information with save.
It doesn’t keep people from stealing our information. The key point is when people steal the information, we knew it.
Suppose I prepare a qubit in a state
or
, and I tell my friend that I store my qubit in a box and go out for lunch. I come back latter, and I know if he could steal my information or not.
To distinguish the qubit in the state
or
secretly, he need to do unitary transformation that preserve my qubit.
He will get his qubit state is
or
But U is unitary so
, thus if
,
=1
He can not steal my information without letting me know it.
Slide31
Quantum Communication
We can use quantum entanglement to send the decode key for classical cryptography.
If Alice want to share with Bob a secret key that they can use latter to decode the information. However they worry about Eva, who can steal their key.
Alice prepare a
entanglement
state
And send one of qubit to Bob
Then Alice and Bob do the measurement on their qubit in Z and X, if they have 100% correlation, they know that they share an entanglement pair of qubit.
Slide32
Quantum Communication
Since Eva can interfere with the system, the state of the three is
But the measurement in Z direction are correlated, thus the state should be
Furthermore, the measurement in X direction are correlated, thus the state should be
So they know that Eva’s state is uncorrelated with their entanglement pairs. They can use their shared entanglement pairs to generate random key latter. Slide33
Conclusion
Quantum computation is promising.
The idea of quantum gate as the driven of time dependent Hamiltonian gives us a hope that we can simulate a quantum system using quantum computer.
Quantum communication is really a big deal.
Slide34
Appendix
Creat
Bell states using
Hadamard
gate