Get an overview of Boulder Opal's functionality and applications

Boulder Opal is a versatile Python toolset that provides everything a research team needs to automate and improve the performance of hardware for quantum computing and quantum sensing.

For any researcher, the challenges of hardware instability, onerous manual tune-up, and increasing complexity with system scaling can dramatically slow—or even totally block—progress. Boulder Opal can help you to overcome these challenges by leveraging five core capabilities based on a powerful discipline called quantum control engineering, which addresses the question, "How can systems that obey the laws of quantum mechanics be efficiently manipulated to create desired behaviors?".

Boulder Opal flexibly delivers these essential capabilities with managed cloud-compute resources; the result is the ability to move faster and solve problems that were otherwise intractable. Boulder Opal is an essential tool for anyone performing theoretical or experimental research in quantum technology.

Capabilities

Characterization

Learning how to operate a quantum device begins with characterizing and identifying key system parameters. This information allows you to build a model of your system that you can use to inform how to design suitable controls to efficiently manipulate your system. These insights also provide operational information such as key coupling rates that can be used in system tuneup and performance optimization. You can check out the hardware characterization topic to learn more.

Design error-robust controls

This is the core task of creating control solutions to manipulate quantum systems so they are resilient to noise, cancel particular Hamiltonian terms, or minimize the duration of an operation. It is the most fundamental incarnation of the control theoretic concept of device optimization—figuring out how best to manipulate a quantum system given a set of available "control knobs". Control design can be performed using either a Hamiltonian model of the system or via closed-loop experimental optimization in the case where a model is incomplete. The control design strategy topic provides a comprehensive overview of the process.

Simulate quantum dynamics

Understanding and anticipating the behavior of a complex quantum device requires simulation under realistic conditions. Boulder Opal's simulation package is designed to address complex dynamics that exceed the capabilities of other packages, handling high-dimensional systems, open and closed-system time-domain dynamics, and easily incorporating various noise and error processes to give you real insights. The tutorials and user guides provide examples of how to use Boulder Opal for simulation.

Automate hardware with AI

Moving rapidly and increasing system complexity requires the replacement of manual tuneup procedures with automated routines that scale efficiently. Closed-loop agents can be directly connected to hardware in order to replace manual parameter scans across entire systems. They can also be used in cases where you have an incomplete system model—the AI agent can learn the underlying dynamics without the operator having to gain specific insight. Useful for both control design in complex systems and hardware calibration. The automation for quantum experiments topic gives a more detailed overview of how Boulder Opal can be used to automate hardware tune-up and optimization.

Verify performance

Ensuring your solutions work well requires analyzing the control solutions to gain insights into their functionality or effectiveness after design. This includes techniques to both probe the expected response and action of the control via simulation under realistic conditions and experimental probing of performance through specific validation routines known to highlight or amplify certain errors. Check out the user guide examples to learn how to use Boulder Opal to evaluate control performance.

Convenience, flexibility, and performance

Flexible high-performance computational graphs

Boulder Opal offers the most flexible way to represent complex research problems using computational graphs. We pioneered their application in quantum control research to empower you to easily tackle a range of problems with a simple and efficient code structure. They also build instant compatibility with TensorFlow to maximize computational performance. In head-to-head comparisons, Boulder Opal can be up to 100x faster than competitive research-grade tools in time to solution, and a unified framework lets you shift seamlessly between computational tasks.

Convenience functionality

Boulder Opal is designed to be both flexible and easy to use. You can use our specialized modules for trapped ions, superconducting qubits, closed-loop optimization, noise reconstruction, and our library of signals to simplify your coding for common tasks. These modules provide pre-built functions and operations to simplify your solution development and implementation, reducing the required amount of code by up to 60%!

Integrations with industry-standard packages

Boulder Opal can be used as a standalone numerical tool or connected directly to hardware so AI agents can automate and accelerate manual tasks. We offer supporting packages to connect to common hardware platforms from Quantum Machines, ArtiQ, and others.

Boulder Opal fully integrates with industry-standard simulation packages like QuTiP, and serves as a powerful extension to expand performance and capability to help researchers address problems that are otherwise out of reach. You can learn how to integrate Boulder Opal with other tools using the provided examples in the user guides.

Performance to unlock new insights

Boulder Opal delivers cloud-based computational acceleration to help researchers accomplish more in less time and with less hassle. Access to managed memory and parallel computing resources in the cloud lets you tackle complex simulations and optimizations on high-dimensional systems that are otherwise impossible without building your own cluster! There's zero user configuration or overhead—just execute simple Python commands and the rest is invisibly managed.

Check out our Get started guide to get set up in just minutes, and quickly learn how to accelerate your research with our hands-on tutorials.