Getting Started

New to Ocean? The following sections describe how to install Ocean tools, what they are and how they fit together, and give examples of using them to solve hard problems on a D-Wave quantum computer.

Initial Set Up

The following steps set up your development environment for Ocean:

1. Installing Ocean Tools

   Installation is not needed if you are using an IDE that implements the Development Containers specification (aka “devcontainers”), whether locally on your system (e.g., VS Code) or cloud-based (e.g., GitHub Codespaces), because you can work in an updated Ocean environment through the Ocean Docker file.

2. Authorizing Access to Leap

   Optionally authorize Ocean to access your Leap account to facilitate token management.

3. Configuring Access to Leap’s Solvers

   Enable the running problems on D-Wave remote compute resources, including quantum-classical hybrid solvers and the D-Wave quantum processing unit (QPU).

Ocean’s Programming Model

Learn Ocean software’s workflow for problem solving.

• Workflow Steps: Formulation and Sampling
• Formulation: Objectives and Models
• Sampling: Minimizing the Objective
• Ocean Software Stack

D-Wave Compute Resources

Use Ocean’s samplers to solve problems on D-Wave’s compute resources (solvers) or locally on your CPU.

• Hybrid Solvers
• Classical Solvers
• Quantum Solvers

Because most industrial problems (large, complex, and hard) are best approached with quantum-classical hybrid solvers, a good place to start is with examples of the End-to-End Examples: Hybrid section. If you wish to learn how to work directly with the quantum computer, see the examples of the End-to-End Examples: QPU section.

End-to-End Examples: Hybrid

For beginners, formulating problems as Quadratic Models can be a more intuitive introduction to solving optimization problems. Nonlinear Models may be more familiar for users with experience in non-linear programming; for many problems, these models—if effectively formulated—are expected to enable superior performance.

Quadratic Models

• Diet Planning solves a mixed-integer linear-programming (MILP) problem using a Leap hybrid CQM solver.

• Bin Packing solves a binary constrained problem using a Leap hybrid CQM solver.

• Stock-Sales Strategy in a Simplified Market solves an integer constrained problem using a Leap hybrid CQM solver.

• Structural Imbalance in a Social Network solves an unconstrained problem using a Leap hybrid BQM solver.

• Large Map Coloring demonstrates using an out-of-the-box Ocean hybrid solver.

• Map Coloring: Hybrid DQM Sampler solves a discrete quadratic model (DQM) using a Leap hybrid DQM solver.

• Problem With Many Variables builds a hybrid workflow and solver for a large graph problem.

Nonlinear Models

These examples solve small instances of known optimization problems using a Leap hybrid nonlinear-model solver.

• Traveling Salesperson helps you start using Leap’s hybrid nonlinear-model solver.

• Vehicle Routing demonstrates more advanced usage options for solving nonlinear models.

End-to-End Examples: QPU

• Vertex Cover solves a small graph problem.

• Constrained Scheduling solves a small constraint satisfaction problem.

• Boolean NOT Gate mathematically formulates a BQM for a two-variable problem.

• Boolean AND Gate demonstrates programming the QPU more directly (minor-embedding).

• Map Coloring example solves a more complex constraint satisfaction problem.

• Multiple-Gate Circuit looks more deeply at minor-embedding.

• Postprocessing with a Greedy Solver improves samples returned from a QPU by post-processing with a classical greedy algorthim.

• Using the Problem Inspector improves minor-embedding on a graph partition problem.

• Working With Different Topologies runs your code on software samplers with different QPU-inspired topologies.

Additional Examples

D-Wave’s dwave-examples GitHub repo contains many more code examples:

• Demos

  Typically in the form of short code examples you can open in a supported cloud-based IDE or copy (clone) locally and run. For example:

  • Nurse scheduling, maze, circuit fault diagnosis, and others provide examples of constraint-satisfaction problems.

  • Map coloring, job-shop scheduling, and others can complement similar examples presented here.

  • RNA folding, portfolio optimization, and others provide prototype applications in a variety of industries.

• Jupyter Notebooks

  These examples, in a web-based interactive environment that includes documentation and code, are helpful for both walking beginners through the theory and practice of problem solving and explaining complex features. They can also serve as a framework in which to develop your own code. For example:

  • Structural imbalance notebook can complement the Structural Imbalance in a Social Network example.

  • Hybrid computing notebooks walk you through using and developing hybrid solvers.

  • Pegasus notebook, reverse annealing notebook, and others explain features of the quantum computer.

Further Learning

• Getting Started with the D-Wave System

  This guide in the System Documentation introduces the D-Wave quantum computer, provides some key background information on how the system works, and explains how to construct a simple problem that the system can solve.

• D-Wave Problem-Solving Handbook

  This guide provides advanced guidance on using D-Wave solvers, in particular QPU solvers. It lists, explains, and demonstrates techniques of problem formulation, minor-embedding, and configuring QPU parameters to optimize performance.

• Package introductions

  The following Ocean packages have extended introductions:

  • The introduction to dimod describes Ocean’s supported models (e.g., BQMs), the format of returned solutions, Symbolic Math, and Scaling for Production.

  • The introduction to dwave-cloud-client discusses how to configure selection of and communications with solvers.

  • The introduction to dwave-hybrid explains how to use the Python framework for running and building hybrid samplers.