Canary Foundation at Stanford · 2015 we announced general availability of the 1000+ qubit D-Wave...

Canary Foundation at Stanford

D-Wave Systems

Murray Thom

February 27th, 2017

Introduction to Quantum Computing


Richard FeynmanRichard FeynmanRichard FeynmanRichard Feynman

1960 1970 1980 1990 2000 2010 2020


Quantum Turing MachineQuantum Turing MachineQuantum Turing MachineQuantum Turing Machine

1950 1960 1970 1980 1990 2000 2010


Quantum Annealing Outlined by Tokyo TechQuantum Annealing Outlined by Tokyo TechQuantum Annealing Outlined by Tokyo TechQuantum Annealing Outlined by Tokyo Tech

PHYSICAL REVIEW E VOLUME 58, NUMBER 5 NOVEMBER 1998

Quantum annealing in the transverse Ising model

Tadashi Kadowaki and Hidetoshi NishimoriDepartment of Physics, Tokyo Institute of Technology, Oh-okayama,

Meguro-ku, Tokyo 152-8551, Japan

(Received 30 April 1998)

We introduce quantum fluctuations into the simulated annealing process of

optimization problems, aiming at faster convergence to the optimal state. Quantum

fluctuations cause transitions between states and thus play the same role as thermal

fluctuations in the conventional approach. The idea is tested by the transverse Ising

model, in which the transverse field is a function of time similar to the temperature in

the conventional method. The goal is to find the ground state of the diagonal part of

the Hamiltonian with high accuracy as quickly as possible. We have solved the time-

dependent Schrödinger equation numerically for small size systems with various

exchange interactions. Comparison with the results of the corresponding classical

(thermal) method reveals that the quantum annealing leads to the ground state with

much larger probability in almost all cases if we use the same annealing schedule.

[S1063-651X~98!02910-9]

1960 1970 1980 1990 2000 2010 2020

https://upload.wikimedia.org/wikipedia/commons/thumb/1/12/Quant-annl.jpg/300px-Quant-annl.jpg


DDDD----Wave Announces 16 Qubit QCWave Announces 16 Qubit QCWave Announces 16 Qubit QCWave Announces 16 Qubit QC

1960 1970 1980 1990 2000 2010 2020

D-Wave Progression

• By 2004 it had become apparent that creating good ideas about quantum computing and looking externally for a research team to use this knowledge to build such a machine wouldn't work. So we decided to do it ourselves. We built our own fabrication facility - a superconducting electronics foundry - to produce the processors required to use quantum effects to compute. We assembled a team of scientists to design, fabricate, and test the processors in our own in-house labs.

• In 2010 we released our first commercial system, the D-Wave One™ quantum computer. We have doubled the number of qubits each 18 months, and in 2013 we shipped our 512-qubit D-Wave Two™ system. In 2015 we announced general availability of the 1000+ qubit D-Wave 2X™ system.

QC Models


Quantum Information ScienceQuantum Information ScienceQuantum Information ScienceQuantum Information Science

Quantum

Computing

Gate Model

Annealing

Topological

One-way/ cluster state

Quantum Cryptography

Quantum key distribution

Quantum Sensor

Quantum information processing

Quantum CommunicationEmerging

Emerging


What is a Quantum Computer? What is a Quantum Computer? What is a Quantum Computer? What is a Quantum Computer?

• Exploits quantum mechanical effects

• Built with “qubits” rather than “bits”

• Operates in an extreme environment

• Enables quantum algorithms to solve

very hard problems

Quantum Processor


Gate Model Quantum ComputingGate Model Quantum ComputingGate Model Quantum ComputingGate Model Quantum Computing

http://www.nature.com/nature/journal/v414/n6866/images/414883a-f1.2.jpg


Quantum Annealing (T=0, N=0)Quantum Annealing (T=0, N=0)Quantum Annealing (T=0, N=0)Quantum Annealing (T=0, N=0)en

ergy

leve

ls

Sol

utio

n

Initi

al s

tate

10 s

HS(t) = (1−s)HI + sHP , s = t/tf


Quantum Annealing (+ Thermal Noise)Quantum Annealing (+ Thermal Noise)Quantum Annealing (+ Thermal Noise)Quantum Annealing (+ Thermal Noise)en

ergy

leve

ls

P0

kBT

System Bath Interaction

10

Dynamical freeze-out

s

SBBS HHtHtH ++= )()(


Topological Quantum ComputingTopological Quantum ComputingTopological Quantum ComputingTopological Quantum Computing

https://static1.squarespace.com/static/51ee6559e4b06fd80f3cb11e/t

/520540fce4b00fb5186e572f/1376076030364/SteveSimon_800_320.jpg

• Microsoft Research

• TU Delft

• And others

Anyons

Non-Abelian Anyons


Photonic/OpticalPhotonic/OpticalPhotonic/OpticalPhotonic/Optical

• University of Bristol

• And others


Trapped IonsTrapped IonsTrapped IonsTrapped Ions

• Oxford

• UofSussex

• UofMD

• Innsbruck

• And others

http://www.nature.com/nature/journal/v464/n7285/images/nature08812-f3.2.jpg


SiliconSiliconSiliconSilicon----based devicesbased devicesbased devicesbased devices

• Intel

• University of New South Wales

• TU Delft

• And others

https://www.engineering.unsw.edu.au/news/quantum-computing-first-two-qubit-logic-gate-in-silicon


Superconducting QubitsSuperconducting QubitsSuperconducting QubitsSuperconducting Qubits

• D-Wave Systems

• Google

• MIT-LL/IARPA

• IBM


DDDD----Wave qubit count has grown exponentiallyWave qubit count has grown exponentiallyWave qubit count has grown exponentiallyWave qubit count has grown exponentially

Qubits(log scale)

‘04 ‘08 ‘12 ‘16

D-Wave One

128

D-Wave Two

512

28

16

4

D-Wave 2X

1000

1

10

100

1,000

10,000

D-Wave 2000Q

2000

‘18‘06 ‘10 ‘14

Next Gen

9

2

Gate

Model

Quantum Annealing


• Space of solutions defines an energy landscape & best solution is lowest valley

• Classical algorithms must walk over this landscape

• Quantum annealing uses quantum effects to go through the mountains

Energy LandscapeEnergy LandscapeEnergy LandscapeEnergy Landscape


Quantum Effects on DQuantum Effects on DQuantum Effects on DQuantum Effects on D----Wave SystemsWave SystemsWave SystemsWave Systems

Superposition

Entanglement

Quantum Tunneling


Quantum Turing MachineQuantum Turing MachineQuantum Turing MachineQuantum Turing Machine

1950 1960 1970 1980 1990 2000 2010


Quantum Annealing Outlined by Tokyo TechQuantum Annealing Outlined by Tokyo TechQuantum Annealing Outlined by Tokyo TechQuantum Annealing Outlined by Tokyo Tech

PHYSICAL REVIEW E VOLUME 58, NUMBER 5 NOVEMBER 1998

Quantum annealing in the transverse Ising model

Tadashi Kadowaki and Hidetoshi NishimoriDepartment of Physics, Tokyo Institute of Technology, Oh-okayama, Meguro-ku, Tokyo 152-

8551, Japan

(Received 30 April 1998)

We introduce quantum fluctuations into the simulated annealing process of optimization problems, aiming at

faster convergence to the optimal state. Quantum fluctuations cause transitions between states and thus play

the same role as thermal fluctuations in the conventional approach. The idea is tested by the transverse Ising

model, in which the transverse field is a function of time similar to the temperature in the conventional method.

The goal is to find the ground state of the diagonal part of the Hamiltonian with high accuracy as quickly as

possible. We have solved the time-dependent Schrödinger equation numerically for small size systems with

various exchange interactions. Comparison with the results of the corresponding classical (thermal) method

reveals that the quantum annealing leads to the ground state with much larger probability in almost all cases if

we use the same annealing schedule.

[S1063-651X~98!02910-9]

1960 1970 1980 1990 2000 2010 2020


Making Use of Quantum StatesMaking Use of Quantum StatesMaking Use of Quantum StatesMaking Use of Quantum States

By Peppergrower - Own work, CC BY-SA 3.0,

https://commons.wikimedia.org/w/index.php?curid=6007495

1

0

Gate

Model


Making Use of Quantum StatesMaking Use of Quantum StatesMaking Use of Quantum StatesMaking Use of Quantum States

1

0

Quantum

Annealing


Quantum Enhanced OptimizationQuantum Enhanced OptimizationQuantum Enhanced OptimizationQuantum Enhanced Optimization

Quantum Hamiltonian is an operator on Hilbert space:

ℋ � = ℰ � ��

�� +��

�

�� + Δ � ��

�

�

Corresponding classical optimization problem:

Obj(��, �� ; ��) =��

�+��

�

��


DDDD----Wave 2X Quantum ProcessorWave 2X Quantum ProcessorWave 2X Quantum ProcessorWave 2X Quantum Processor

Qubits within red boxes

Overview of D-Wave


First and only commercial quantum computerFirst and only commercial quantum computerFirst and only commercial quantum computerFirst and only commercial quantum computer

• Customers include Google, NASA, Lockheed, University of

Southern California, Los Alamos National Laboratory, Temporal

Defense Systems

• 150 U.S. patents

• 160 employees, 45 with Ph.D.

• HQ in Vancouver, B.C.

• Founded in 1999


MissionMissionMissionMission

To help solve the most challenging problems in the multiverse:

• Optimization

• Machine Learning

• Monte Carlo/Sampling


Better Answers for Hard ProblemsBetter Answers for Hard ProblemsBetter Answers for Hard ProblemsBetter Answers for Hard Problems

Graph

Coloring

Factoring

V&V

Constraint Satisfaction

Monte

Carlo

Financial

Modeling

Filtering

Sampling

Scaling Error

Treatment

Topology

Quantum Research

Optimization/Decision Support

Scheduling Logistics

Planning

Deep Learning

Structured

Prediction

Boltzmann

Machines


Google Optimization BenchmarksGoogle Optimization BenchmarksGoogle Optimization BenchmarksGoogle Optimization Benchmarks (2013)(2013)(2013)(2013)

0.001

0.01

0.1

1

10

100

1000

10000

0 100 200 300 400 500

Me

dia

n t

ime

to

be

st s

olu

tio

n (

s)

Problem size (number of qubits)

Series1

Series2

Series3

Series4

11000 x

Timing Benchmark – Smaller is Better

11,000xD-WAVE II


Machine Learning: Binary ClassificationMachine Learning: Binary ClassificationMachine Learning: Binary ClassificationMachine Learning: Binary Classification

• Traditional algorithm recognized car about 84% of the time

• Google/D-Wave Qboostalgorithm implemented to recognize a car (cars have big shadows!)

• “Quantum Classifier” was more accurate (94%) and more efficient

• Ported quantum classifier back to traditional computer, more accurate and fewer CPU cycles (less power)!


Google Blog December 8, 2015Google Blog December 8, 2015Google Blog December 8, 2015Google Blog December 8, 2015http://googleresearch.blogspot.ca/2015/12/when-can-quantum-annealing-win.htmlWhen can Quantum Annealing win?

Tuesday, December 08, 2015Posted by Hartmut Neven, Director of Engineering-

During the last two years, the Google Quantum AI team has made progress in understanding the physics governing quantum annealers. We recently applied these new insights to construct proof-of-principle optimization problems and programmed these into the D-Wave 2X quantum annealer that Google operates jointly with NASA. The problems were designed to demonstrate that quantum annealing can offer runtime advantages for hard optimization problems characterized by rugged energy landscapesWe found that for problem instances involving nearly 1000 binary variables, quantum annealing significantly outperforms its classical counterpart, simulated annealing. It is more than 108 times faster than simulated annealing running on a single core.

100,000,000x

Quantum Computing System


DDDD----Wave Container Wave Container Wave Container Wave Container ----““““SCIFSCIFSCIFSCIF----like” like” like” like” ---- No No No No RFRFRFRF InterferenceInterferenceInterferenceInterference


System ShieldingSystem ShieldingSystem ShieldingSystem Shielding

• 16 Layers between the quantum chip

and the outside world

• Shielding preserves the quantum

calculation


Processor EnvironmentProcessor EnvironmentProcessor EnvironmentProcessor Environment

• Cooled to 0.015 Kelvin, 175x colder

than interstellar space

• Shielded to 50,000× less than Earth’s

magnetic field

• In a high vacuum: pressure is 10 billion

times lower than atmospheric pressure

• On low vibration floor

• <25 kW total power consumption – for

the next few generations

15mK


DDDD----Wave Wave Wave Wave 2000Q 2000Q 2000Q 2000Q Quantum ProcessorQuantum ProcessorQuantum ProcessorQuantum Processor


Processing Using DProcessing Using DProcessing Using DProcessing Using D----Wave Wave Wave Wave

• A lattice of superconducting loops (qubits)

• Chilled near absolute zero to quiet noise

• User maps a problem into search for “lowest point in a vast landscape” which corresponds to the best possible outcome

• Processor considers all possibilities simultaneously to satisfy the network of relationships with the lowest energy

• The final state of the qubits yields the answer


�

��

��

Qubit assuperconducting loop

The EnergyPotential

The QubitThe QubitThe QubitThe Qubit


The The The The CouplingCouplingCouplingCoupling


QubistQubistQubistQubist


Colors encoded in unit cellsColors encoded in unit cellsColors encoded in unit cellsColors encoded in unit cells


“Virtual” QUBO

qbsolv

DDDD----Wave Software EnvironmentWave Software EnvironmentWave Software EnvironmentWave Software Environment

LANL Assembler

Environment/

Libraries

1Qbit SDK

JADE/QuellE…

QUBOINTERMEDIATE

REPRESENTATION

TARGET SYSTEM

HOST LIBRARY AND COMMAND

LINE INTERFACE

C, C++, MATLAB Python

DW

SAPI SYSTEM INTERFACE AND

CONTROL

TRANSLATORSQSAGE

OptimizationConstraint

Satisfaction

ToQ

SamplingSAT, ML

?� � �

QUANTUM MACHINE INSTRUCTIONQMI

APPLICATIONS

“QUORTRAN” COMPILERS“Q++”

PRODUCT PROTOTYPE CONCEPT

Canary Foundation at Stanford · 2015 we announced general availability of the 1000+ qubit D-Wave...

Documents

Transcript of Canary Foundation at Stanford · 2015 we announced general availability of the 1000+ qubit D-Wave...