noisy intermediate-scale quantum computing: a post …...noisy intermediate-scale quantum computing:...
Post on 22-Jul-2020
15 Views
Preview:
TRANSCRIPT
Noisy intermediate-scale quantum computing: A post-quantum supremacy perspective
Omar ShehabIonQ/UMD
27 January 2020shehab@ionq.com
Quantum Information Seminar Series | The University of Tennessee, Knoxville
1
www.ionq.coIONQ Proprietary Material
Collaborators
2
IonQ● Matthew Kissan● Yunseong Nam
ORNL● Raphael Pooser● Eugene Dumitrescu● Alex McCaskey● Pavel Lougovski● Thomas Papenbrock
UMD● Kevin Landsman● Norbert Linke● Cinthia Huerta
Alderete● Nhung H. Nguyen● Daiwei Zhu● Chris Monroe
Stanford● Isaac H Kim
JHU/APL● Greg Quiroz
3
Outline
● What is quantum computing?● Building trapped-ion quantum computers● Near-term quantum computing● An example of NISQ algorithm● Optimizing NISQ algorithms● Post-quantum supremacy era
3
4
What is quantum computing?
● Proposed by Feynman in 1985 with the motivation of simulating quantum systems
● Based on the principles of quantum mechanics
● Potential advantage over classical computing comes from quantum phenomena like quantum superposition, quantum entanglement, and quantum measurement
IonQ
IBM
Rigetti4
5
What is quantum computing?
● Not a silver bullet for all computationally hard (for example NP-complete) problems
● Non-von Neumann architecture● Data is stored in qubits residing in Hilbert space● Logical operations are complex unitary matrices● Readout is probabilistic● Algorithms are often visualized as “quantum
circuits”
5
6
What is quantum computing?
Example: Quantum algorithm to find an element from an unstructured database.
Source: Wikipedia page of the Grover’s algorithm 6
7
P
P
Quantum computational complexity
PP
BQPNP
NP-complete
QMA
PSPACEn X n chess/Go
Traveling salesperson
Graph isomorphism
Integer factoring
Primality checking
Finding the ground state of k-local Hamiltonian
Extending the graphics from Scott Aaronson’s lecture notes 7
8
Quantum algorithm timeline
1985 1994 1996 2005 2008 20152014
Feynman proposes quantum computing
Algorithm forprime factoringWith exponentialspeedup
Unstructured database search with quadratic speedup
Quantum chemistrySimulation withExponential speedup
Solving special systems of linear equations with exponential speedup
Non-convex optimization with Comparable approximation ratio
Quantum chemistry simulation onnear-term quantumcomputers
* References in the back up slides8
9
Quantum algorithm timeline
1985 1994 1996 2005 2008 20152014
Feynman proposes quantum computing
Algorithm forprime factoringWith exponentialspeedup
Unstructured database search with quadratic speedup
Quantum chemistrySimulation withExponential speedup
Solving special systems of linear equations with exponential speedup
Non-convex optimization with Comparable approximation ratio
Quantum chemistry simulation onnear-term quantumcomputers
* References in the back up slides
Triggered funded academic research programs
9
10
Quantum algorithm timeline
1985 1994 1996 2005 2008 20152014
Feynman proposes quantum computing
Algorithm forprime factoringWith exponentialspeedup
Unstructured database search with quadratic speedup
Quantum chemistrySimulation withExponential speedup
Solving special systems of linear equations with exponential speedup
Non-convex optimization with Comparable approximation ratio
Quantum chemistry simulation onnear-term quantumcomputers
* References in the back up slides
Triggered private industrial effort
mainly focusing onchemistry/physics
simulation
10
11
Quantum algorithm timeline
1985 1994 1996 2005 2008 20152014
Feynman proposes quantum computing
Algorithm forprime factoringWith exponentialspeedup
Unstructured database search with quadratic speedup
Quantum chemistrySimulation withExponential speedup
Solving special systems of linear equations with exponential speedup
Non-convex optimization with Comparable approximation ratio
Quantum chemistry simulation onnear-term quantumcomputers
* References in the back up slides
Potential for commercial
incentives still not clear for
ML/DL/optimization
11
12
Potential market
12
Pharma Energy Transportation Financemolecular design
drug deliverymarket models
optimizationchemical processes
material designnavigation
logistics
Quantum computing @ IonQ
14
Coming soon to a cloud near you!
Microsoft Ignite, Nov 4 201914
AWS Invent, Dec 4 2019
15
Building quantum computers @ IonQ
Source: arXiv:1903.0818115
16
The quantum processing unit
Linear computation: montage of a photo of the chip containing the trapped ions and an image of the ions in a 1D array (Courtesy: Christopher Monroe) 16
APS March MeetingLos Angeles, CA, USA, March 7th, 2018
Trapped Ion Hyperfine Qubit: 171Yb+
2P1/22.1 GHz
γ/2π = 20 MHz
Single ion
2S1/2
|↓〉 = |0,0〉
|↑〉 = |1,0〉
• State-dependent fluorescence provides high fidelity detection
S. Olmschenk et al., PRA 76, 052314, Slide (2007)From the March Meeting talk of Jungsang Kim
369 nm
(811.9THz)
APS March MeetingLos Angeles, CA, USA, March 7th, 2018
~5 μm↑
↓
↑
↓ δ
r
Cirac and Zoller (1995)Mølmer & Sørensen (1999)
Solano, de Matos Filho, Zagury (1999)Milburn, Schneider, James (2000)
From the March Meeting talk of Jungsang Kim
δ ~ 10 nmeδ ~ 500 Debye
dipole-dipole coupling
for full entanglement
Native Ion Trap Operation: “Ising” gate
Tgate ~ 10−100 μs F ~ 98% – 99.9%
Entangling Trapped Ion Qubits
19
Assemblying large number of qubits
19More ways to apply fundamental logical operation directly on hardware
20
Platform comparison
20
QC Platform Qubit Type T1 / T2 (s) 1 & 2 QubitGate Error
#2-Q GatesRun in a Circuit
Maximum# Qubits
Entangled
Trapped Ions Natural > 1010 / 103 < 10-5 / 10-3 56 20
Neutral Atoms Natural < 10-4 / 10-4 < 10-3 / 0.03 1 2
Photonic Circuits Engineered Photon Loss Probabilistic - -
Superconductors Engineered < 10-3 / 10-3 < 10-3 / 0.005 15 9
Silicon Spin Engineered 10-3/10-5 > 10-3/0.1 1 2
21
Nature’s qubit
21
QC Platform Qubit Type T1 / T2 (s) 1 & 2 QubitGate Error
#2-Q GatesRun in a Circuit
Maximum# Qubits
Entangled
Trapped Ions Natural > 1010 / 103 < 10-5 / 10-3 56 20
Neutral Atoms Natural < 10-4 / 10-4 < 10-3 / 0.03 1 2
Photonic Circuits Engineered Photon Loss Probabilistic - -
Superconductors Engineered < 10-3 / 10-3 < 10-3 / 0.005 15 9
Silicon Spin Engineered 10-3/10-5 > 10-3/0.1 1 2
22
Time available to finish computation
22
QC Platform Qubit Type T1 / T2 (s) 1 & 2 QubitGate Error
#2-Q GatesRun in a Circuit
Maximum# Qubits
Entangled
Trapped Ions Natural > 1010 / 103 < 10-5 / 10-3 56 20
Neutral Atoms Natural < 10-4 / 10-4 < 10-3 / 0.03 1 2
Photonic Circuits Engineered Photon Loss Probabilistic - -
Superconductors Engineered < 10-3 / 10-3 < 10-3 / 0.005 15 9
Silicon Spin Engineered 10-3/10-5 > 10-3/0.1 1 2
23
Noise rate
23
QC Platform Qubit Type T1 / T2 (s) 1 & 2 QubitGate Error
#2-Q GatesRun in a Circuit
Maximum# Qubits
Entangled
Trapped Ions Natural > 1010 / 103 < 10-5 / 10-3 56 20
Neutral Atoms Natural < 10-4 / 10-4 < 10-3 / 0.03 1 2
Photonic Circuits Engineered Photon Loss Probabilistic - -
Superconductors Engineered < 10-3 / 10-3 < 10-3 / 0.005 15 9
Silicon Spin Engineered 10-3/10-5 > 10-3/0.1 1 2
24
Longest sequence of logical operation
24
QC Platform Qubit Type T1 / T2 (s) 1 & 2 QubitGate Error
#2-Q GatesRun in a Circuit
Maximum# Qubits
Entangled
Trapped Ions Natural > 1010 / 103 < 10-5 / 10-3 56 20
Neutral Atoms Natural < 10-4 / 10-4 < 10-3 / 0.03 1 2
Photonic Circuits Engineered Photon Loss Probabilistic - -
Superconductors Engineered < 10-3 / 10-3 < 10-3 / 0.005 15 9
Silicon Spin Engineered 10-3/10-5 > 10-3/0.1 1 2
25
Largest input size used so far
25
QC Platform Qubit Type T1 / T2 (s) 1 & 2 QubitGate Error
#2-Q GatesRun in a Circuit
Maximum# Qubits
Entangled
Trapped Ions Natural > 1010 / 103 < 10-5 / 10-3 56 20
Neutral Atoms Natural < 10-4 / 10-4 < 10-3 / 0.03 1 2
Photonic Circuits Engineered Photon Loss Probabilistic - -
Superconductors Engineered < 10-3 / 10-3 < 10-3 / 0.005 15 9
Silicon Spin Engineered 10-3/10-5 > 10-3/0.1 1 2
26
History of improvement for gate error
26
27
Ions vs. Superconductors
27
F FFFF
F : Failure
1.00
0.75
0.50
0.25
0.00
BV4 BV10 HS2 HS4 Toffoli Fredkin Or Peres QFT Adder
IBMQ5 IBMQ14 IBMQ16 Rigetti Trapped Ions
MeasuredSuccess
Rate
P. Murali, et al., ISCA 2019 (arXiv:1905.11349)
Benchmark Algorithm
28
Scaling technology
28
Modular shuttling between multiple zones
Shuttling of ion chain: tape+head model
Modular photonic connections between chips
29
The promise
29
SIZE OF PROBLEM (BITS)
TIM
E T
O S
OLU
TIO
N1 BILLION YRS
1 MILLION YRS
1,000 YRS
10 YRS
1 YR
100 YRS
1 DAY
1 HR
100 SEC
1 SEC
1 MO
QUANTUM COMPUTER
CLASSICAL COMPUTER: STATE OF THE ART
CLASSICAL COMPUTER: 30 YEARS FROM NOW
(MOORE’S LAW)
Classical computers are a dead end
for certain classes of problems.
Factoring Problem
30
The promise
30
SIZE OF PROBLEM (BITS)
TIM
E T
O S
OLU
TIO
N1 BILLION YRS
1 MILLION YRS
1,000 YRS
10 YRS
1 YR
100 YRS
1 DAY
1 HR
100 SEC
1 SEC
1 MO
QUANTUM COMPUTER
CLASSICAL COMPUTER: STATE OF THE ART
CLASSICAL COMPUTER: 30 YEARS FROM NOW
(MOORE’S LAW)
So called “quantum supremacy”
Factoring Problem
Near-term quantum computing
32
The roadmap (near term)
32
Algorithm Improvements
Log(
Perf
orm
ance
)C
ompu
tatio
nal P
ower
Quantum Machine Learning
Factoring
Quantum Chemistry/Simulations
2015 2017 2019 2021 2023 2025
H2HeH
H2OO3
Nitrogenase
Simple Organics
33
The roadmap (near term)
33
Algorithm Improvements
Log(
Perf
orm
ance
)C
ompu
tatio
nal P
ower
Quantum Machine Learning
Factoring
Quantum Chemistry/Simulations
2015 2017 2019 2021 2023 2025
H2HeH
H2OO3
Nitrogenase
Simple Organics
Published works
34
The roadmap (near term)
34
Algorithm Improvements
Log(
Perf
orm
ance
)C
ompu
tatio
nal P
ower
Quantum Machine Learning
Factoring
Quantum Chemistry/Simulations
2015 2017 2019 2021 2023 2025
H2HeH
H2OO3
Nitrogenase
Simple Organics
35
Practical challenges of building a QC
● Computation makes qubits (unit of quantum registers) interact with each other
● Qubits are susceptible to noise coming from the surrounding environment
● It is hard to filter out unwanted interaction while allowing computationally meaningful interactions
Source: Preskill, 201835
36
Practical challenges of building a QC
GPU errors (Tiwari et. al 2015) Quantum processing unit (QPU) errors (Wright et. al 2019)
36
37
NISQ computing
● Noisy intermediate-scale quantum (NISQ) computers = First generation quantum computers
● Name coined by John Preskill (2018)● Amount of computation very limited due to high
failure rate● Upper bound for quantum advantage is quadratic
when the structure of the problem is not known● With knowledge about the structure it may
improve
37
38
Algorithmic framework on NISQ computer
● The computational problem is encoded in a quantum-native format
● The goal is to construct a quantum state which encodes the solution
● This construction is done variationally using a parametrized quantum circuit (= a simple quantum algorithm with lots of variables)
● For many problems, the solution is the ground state of the system
● Classical stochastic optimization routines are used to find the ground state
38
39
Domains of applications
● Simulation of physics/chemistry● Hard optimization problems
39
40
Standard NISQ algorithms
● Simulation of physics/chemistry○ Quantum advantage is well understood
● Hard optimization problems○ Quantum advantage is still under investigation
40
41
NISQ algorithmic schemes
● Simulation of physics/chemistry○ Variational quantum eigensolver algorithm (McClean et. al 2015)
● Hard optimization problems○ Quantum approximate optimization algorithm (Farhi et. al 2014)
41
42
Quantum approximate optimization algorithm
● Define a measurable quantum quantity which encodes the solution to the problem
● Design a tunable method to explore the quantum search space (aka run a quantum algorithm)
● Keep tuning the method using a classical routine until you get a reasonable solution
42
NISQ algorithm example
44
An example NISQ problem
● Compute the MAXCUT of the following graph
44
1 2
34
45
An example NISQ problem
● Compute the MAXCUT of the following graph
45
1 2
34
● General MAXCUT problem is NP-complete
● Application: Pattern recognition, electronic circuitry design, state problems in statistical physics, and combinatorial optimization
46
An example NISQ problem
● Compute the MAXCUT of the following graph
46
1 2
34
● Best known exact algorithm needs exponential time
● Approximation algorithms are used in practice
47
An example NISQ problem
● Compute the MAXCUT of the following graph
47
1 2
34
● MAXCUT = 4● Two solutions
48
An example NISQ problem
● Let’s label the two partitions as 0 and 1.● 0000 => all vertices belong to partition #0.● 1010 => the first and third nodes belong to partition
#1.● There are 24 = 16 possible partitions.● A search problem!
48
49
An example NISQ problem
● A classical exact algorithm searches the solution in the integer space
● A weighted version of the problem means searching in the real space
● A quantum search (=optimization algorithm) algorithm looks for the solution in a complex vector space which is huuuuuuge!
49
50
An example NISQ problem
● A classical exact algorithm searches the solution in the integer space
● A weighted version of the problem means searching in the real space
● A quantum search (=optimization algorithm) algorithm looks for the solution in a complex vector space which is huuuuuuge!
50
Good or bad?
51
The Grover speed up
● Thanks to unique quantum phenomenon - destructive interference while exploring the search tree
● Unstructured search may speed up quadratically● Theoretical upper limit● In practice, fractional Grover speedup is all you
can hope
51
52
Back to the example
● Translate the problem into quantum-friendly language
● For any edge <j,k> betweek the nodes j and k, define the following:
C<j,k>
= ½(-σzj σz
k + 1)
52
1 2
34
53
An example NISQ problem
● Translate the problem into quantum-friendly language
● For any edge <j,k> betweek the nodes j and k, define the following:
C<j,k>
= ½(-σzj σz
k + 1)
53
● σz = ● The lowest eigenvalue occurs
when the nodes j and k are both in the same partition
● The highest eigenvalue occurs when the nodes j and k are in different partitions
54
Tl;dr: Still a hard problem
● C = ∑j,k
C<j,k>
gives you the CUT● C is to be measured on a quantum system● Maximizing C or minimizing -C gives you the
MAXCUT● Minimizing -C = Sending the quantum system
encoding the problem to lowest energy state● Finding the lowest energy of a system is NP-hard
in classical physics● It is QMA-hard in quantum physics
54
55
An example NISQ problem
● For each edge, vary the probability of the connected nodes being in different or same partition
● Find a combination of these probabilities which yields the lowest value for -C
55
56
An example NISQ problem
● Varying the probability of the connected nodes = switching the quantum system from one state to another
● Hopefully one of those states will be the (approximate?) solution
56
1 2
This angle dictates the probability
57
An example NISQ problem
57
58
An example NISQ problem
58
Creates an initial superposition of complete search space
59
An example NISQ problem
59
Corresponds to the edge between node #1 and #2
60
An example NISQ problem
60
61
An example NISQ problem
61
62
An example NISQ problem
62
63
An example NISQ problem
63
64
An example NISQ problem
64
Checks symmetry
65
An example NISQ problem
● For each term in C, measure the qubit to determine the probability of being in 1
● Variationally update γ and β to get to the lowest energy
● Repetition of this layer can be viewed as constrained optimization problem which yields higher approximation ratio
● End result will be close to -4 but never beyond that
65
Optimizing NISQ algorithms
67
Past causal cones?
Prof. Glen Evenbly*Prof. Guifré Vidal
Evenbly-Vidal 2013
68
A “harder” problem
68Shehab et. al 2019
69
A “harder” problem
69Shehab et. al 2019
70
Improvements by noise reduction?
70Shehab et. al 2019
71
Improvements by noise reduction?
71Shehab et. al 2019
● 40% increase in accurary● 78% fewer measurements
72
Solving practical problems?
● Practical problems may contain millions of nodes and edges● If one node = one qubit, we need millions of qubits● Millions of edges ≈ 1/million noise rate in each logical operation● Theoretical upper limit on quantum advantage is quadratic speedup● Special knowledge about the structure of the problem might help go
beyond that● Loading classical input onto a quantum computer is not very efficient
72
73
When a problem is quantum-worthy?
73Randomly chosen instances by Zhou, 2003
74
Informed hybrid approach?
74
● Practical NP-complete problems are often easy● Hard to get the exact solution● Think about a search problem with gradient descent where most of the
search landscape is smooth but the basin of the global minima is rugged● Classical computer should handle most of it and delegate to quantum
when the solution landscape is close to random
75
The big picture
Workshop on Theory of Deep Learning: Where next? IAS, Princeton, NJ 2019 Workshop on Quantum Machine Learning, 2018
75
76
Rays of hope
● HHL algorithm with exponential speedup for systems of linear equation (Harrow et. al, 2009)
● Quadratic speedup for SAT problems (Dunjko et. al, 2018)● Quantum gradient descent (Kerenidis 2019)● Quadratic speedup for semidefinite programming solvers (Brandão
et. al, 2019)● Error reduction with past causal cones for non-convex quantum
optimization (Shehab et. al 2019)● Variational search space reduction with chaos theory for non-convex
quantum optimization (Shehab et. al anytime soon at your nearest arXiv)
76
The era of post-quantum supremacy
78
How it all started?
78
79
Formalization of the experiment
79
80
The Google result
80
81
The IBM response
81
82
What is the point?
82
● Amazing progress in the hardware
Google (The supremacy paper) IBM (Jay Gambetta)
83
What is the point?
83
● New ideas in theoretical computer science
84
Thank you and we are hiring!
top related