book

Programming Quantum Computers

by Eric R. Johnston, Nic Harrigan, Mercedes Gimeno-Segovia

July 2019

Intermediate to advanced

333 pages

8h 13m

English

O'Reilly Media, Inc.

Read now

Unlock full access

How This Book Is StructuredConventions Used in This BookUsing Code ExamplesO’Reilly Online LearningHow to Contact UsAcknowledgments
Required BackgroundWhat Is a QPU?A Hands-on ApproachA QCEngine PrimerNative QPU InstructionsSimulator LimitationsHardware LimitationsQPU Versus GPU: Some Common Characteristics
A Quick Look at a Physical QubitIntroducing Circle NotationCircle SizeCircle RotationThe First Few QPU OperationsQPU Instruction: NOTQPU Instruction: HADQPU Instruction: READQPU Instruction: WRITEHands-on: A Perfectly Random BitQPU Instruction: PHASE(θ)QPU Instructions: ROTX(θ) and ROTY(θ)COPY: The Missing OperationCombining QPU OperationsQPU Instruction: ROOT-of-NOTHands-on: Quantum Spy HunterConclusion
Circle Notation for Multi-Qubit RegistersDrawing a Multi-Qubit RegisterSingle-Qubit Operations in Multi-Qubit RegistersReading a Qubit in a Multi-Qubit RegisterVisualizing Larger Numbers of QubitsQPU Instruction: CNOTHands-on: Using Bell Pairs for Shared RandomnessQPU Instructions: CPHASE and CZQPU Trick: Phase KickbackQPU Instruction: CCNOT (Toffoli)QPU Instructions: SWAP and CSWAPThe Swap TestConstructing Any Conditional OperationHands-on: Remote-Controlled RandomnessConclusion
Hands-on: Let’s Teleport SomethingProgram WalkthroughStep 1: Create an Entangled PairStep 2: Prepare the PayloadStep 3.1: Link the Payload to the Entangled PairStep 3.2: Put the Payload into a SuperpositionStep 3.3: READ Both of Alice’s QubitsStep 4: Receive and TransformStep 5: Verify the ResultInterpreting the ResultsHow Is Teleportation Actually Used?Fun with Famous Teleporter Accidents
Strangely DifferentArithmetic on a QPUHands-on: Building Increment and Decrement OperatorsAdding Two Quantum IntegersNegative IntegersHands-on: More Complicated MathGetting Really QuantumQuantum-Conditional ExecutionPhase-Encoded ResultsReversibility and Scratch QubitsUncomputingMapping Boolean Logic to QPU OperationsBasic Quantum LogicConclusion
Hands-on: Converting Between Phase and MagnitudeThe Amplitude Amplification IterationMore Iterations?Multiple Flipped EntriesUsing Amplitude AmplificationAA and QFT as Sum EstimationSpeeding Up Conventional Algorithms with AAInside the QPUThe IntuitionConclusion
Hidden PatternsThe QFT, DFT, and FFTFrequencies in a QPU RegisterThe DFTReal and Complex DFT InputsDFT EverythingUsing the QFTThe QFT Is FastInside the QPUThe IntuitionOperation by OperationConclusion

Learning About QPU OperationsEigenphases Teach Us Something UsefulWhat Phase Estimation DoesHow to Use Phase EstimationInputsOutputsThe Fine PrintChoosing the Size of the Output RegisterComplexityConditional OperationsPhase Estimation in PracticeInside the QPUThe IntuitionOperation by OperationConclusion
Noninteger DataQRAMVector EncodingsLimitations of Amplitude EncodingAmplitude Encoding and Circle NotationMatrix EncodingsHow Can a QPU Operation Represent a Matrix?Quantum Simulation
Phase LogicBuilding Elementary Phase-Logic OperationsBuilding Complex Phase-Logic StatementsSolving Logic PuzzlesOf Kittens and TigersGeneral Recipe for Solving Boolean Satisfiability ProblemsHands-on: A Satisfiable 3-SAT ProblemHands-on: An Unsatisfiable 3-SAT ProblemSpeeding Up Conventional Algorithms
What Can a QPU Do for Computer Graphics?Conventional SupersamplingHands-on: Computing Phase-Encoded ImagesA QPU Pixel ShaderUsing PHASE to DrawDrawing CurvesSampling Phase-Encoded ImagesA More Interesting ImageSupersamplingQSS Versus Conventional Monte Carlo SamplingHow QSS WorksAdding ColorConclusion
Hands-on: Using Shor on a QPUWhat Shor’s Algorithm DoesDo We Need a QPU at All?The Quantum ApproachStep by Step: Factoring the Number 15Step 1: Initialize QPU RegistersStep 2: Expand into Quantum SuperpositionStep 3: Conditional Multiply-by-2Step 4: Conditional Multipy-by-4Step 5: Quantum Fourier TransformStep 6: Read the Quantum ResultStep 7: Digital LogicStep 8: Check the ResultThe Fine PrintComputing the ModulusTime Versus SpaceCoprimes Other Than 2
Solving Systems of Linear EquationsDescribing and Solving Systems of Linear EquationsSolving Linear Equations with a QPUQuantum Principle Component AnalysisConventional Principal Component AnalysisPCA with a QPUQuantum Support Vector MachinesConventional Support Vector MachinesSVM with a QPUOther Machine Learning Applications
From Circle Notation to Complex VectorsSome Subtleties and Notes on TerminologyMeasurement BasisGate Decompositions and CompilationGate TeleportationQPU Hall of FameThe Race: Quantum Versus Conventional ComputersA Note on Oracle-Based AlgorithmsDeutsch-JozsaBernstein-VaziraniSimonQuantum Programming LanguagesThe Promise of Quantum SimulationError Correction and NISQ DevicesWhere Next?BooksLecture NotesOnline Resources

Content preview from Programming Quantum Computers

Chapter 11. Quantum Supersampling

From pixel-based adventure games to photorealistic movie effects, computer graphics has a history of being at the forefront of computing innovations. Quantum Image Processing (QIP) employs a QPU to enhance our image processing capabilities. Although very much in its infancy, QIP already offers exciting examples of how QPUs might impact the field of computer graphics.

What Can a QPU Do for Computer Graphics?

In this chapter we explore one particular QIP application called Quantum Supersampling (QSS). QSS leverages a QPU to find intriguing improvements to the computer graphics task of supersampling, as is summarized in Figure 11-1. Supersampling is a conventional computer graphics technique whereby an image generated by a computer at high resolution is reduced to a lower-resolution image by selectively sampling pixels. Supersampling is an important step in the process of producing usable output graphics from computer-generated images.

QSS was originally developed as a way to speed up supersampling with a QPU, and by this metric it failed. However, numerical analysis of the results revealed something interesting. Although the final image quality of QSS (measured as error per pixel) is about the same as for existing conventional methods, the output image has a different kind of advantage.