Boulder Opal is a versatile toolkit that provides everything you need to improve and automate the performance of hardware for quantum computing and quantum sensing. The core capabilities of Boulder Opal fall under the framework of quantum control—the discipline that addresses the question, "How can systems that obey the laws of quantum mechanics be efficiently manipulated to create desired behaviors?"
The quantum control functionality underlying Boulder Opal comprises three core activities: System Identification, Control Design, and Performance Verification.
- System identification: involves characterizing key system parameters such as energy levels and interactions (AKA Hamiltonian parameters and sources of performance degradation essential for improving hardware). This is generally accomplished by probing the quantum system and performing analysis on the resulting data. For instance, you may choose to characterize the transmission lines in a hardware system using a qubit as a probe, or perform noise spectroscopy on an arbitrary Hamiltonian term. This information together results in a model of the system which can be used to inform how to design suitable controls for your system in the next activity. You can learn more about the process of characterizing hardware with measurements and data fusion in another topic.
- Control design: otherwise known as control optimization, is the creation of control solutions to manipulate quantum systems so they are resilient to noise, to cancel particular Hamiltonian terms, or to minimize control duration. This can be performed using either a Hamiltonian model of the system or via closed-loop experimental interrogation, and will generally involve a critical calibration step on the hardware. Boulder Opal favors robust control which delivers high performance even in the presence of imperfections in your system, and offers multiple techniques to achieve this, such as stochastic optimization. Boulder Opal also offers various approaches to incorporate realistic hardware constraints, by building solutions using smooth basis functions, accounting for nonlinear couplings, directly adding band limits and symmetry requirements on controls. A full accounting of the optimization capabilities is available from the Boulder Opal user guides, and you can review a comprehensive overview of the process of choosing a control design strategy in another topic.
- Performance verification: involves analyzing control solutions to gain insights into their functionality or effectiveness. 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 for easy tuneup. A wide range of simulation capabilities are available and described in the Boulder Opal user guides.
Boulder Opal is designed to be both flexible and convenient to use. You can use a wide range of convenience functions throughout and take advantage of our expert selection of the best performing solutions for critical features like the choice of optimization strategy. Browsing the Boulder Opal user guides will give you insight into the convenience capabilities on offer.
For those users seeking maximum flexibility and customization, our tools expose an intuitive representation for all of the tasks outlined above based on dataflow graphs as the core representation of quantum systems and associated workflows. All key activities - from system identification through to control design and performance simulation can be accommodated in this framework.
Underlying the functionality of Boulder Opal is a powerful range of tools for automation that allows you to leverage AI to accelerate various tasks in your research.
Above, you read that Boulder Opal permits model-free control optimization, but this is only one example of how Boulder Opal's powerful AI agents can be deployed. You can apply Boulder Opal's closed-loop optimization tools to a variety of tasks from microwave or laser hardware calibration to PID tuneup. Boulder Opal also offers a highly capable learning control toolset based on Deep Reinforcement Learning which can be used in a broad range of tasks. Overall, Boulder Opal provides a comprehensive and highly configurable toolset for hardware automation in quantum experiments.
In leveraging these automation techniques directly in hardware, it's easy to seamlessly integrate the outputs of Boulder Opal with any experimental system via our powerful python interface. In addition you can deploy our convenience packages allowing simple integration of Boulder Opal software or output data with specific control electronic platforms .
Our how-to user guides provide detailed step-by-step instructions on how to solve real-world problems using Boulder Opal at a granular level. We've added labels to help you quickly understand the relationship between different user guides and to the overall workflow above. Each user guide provides both a summary workflow to follow and also a worked example containing fully executable code.
In addition we provide Application Notes which highlight new research results realized through use of a range of Boulder Opal capabilities. These push the limits of published literature on quantum control and also demonstrate how flexibly incorporating different parts of the Boulder Opal toolset can solve major control challenges in the field. Finally, we also provide full technical reference documentation.
Now that you're ready, check out our Get started guide for installation and initialization instructions and you'll be on your way!