Orquestra® Questions Answered
Orquestra® Questions Answered
It was great to have the opportunity to give more information about Orquestra® and its commercial release. As promised, in this blog I’m going to answer all of the questions I couldn’t get to during the live webinar in early October, and I’ve also put in some others that I think are most useful for understanding Orquestra or clearing up misconceptions.
In case you didn’t get to attend, you can check out the webinar here.
The first point to note is that Orquestra doesn’t replace SDKs and libraries like Qiskit, it automates the management of them. There are many tasks that can be done in Orquestra that can’t be done with just Qiskit including
For more info on the benefits of workflows for quantum computing, give this blog a read!
Depending on the needs of your organization, we offer different packages and engagement modalities. First, we can provide a license to Orquestra, with a variety of support models that leverages Zapata’s world-class quantum scientists to enable your efforts. A second option is focused around consulting that can lead to a proof of concept. Here we recommend a conversation around 5-10 use-cases of interest. We can have our technical team evaluate these use-cases and provide an evaluation at a face-to-face (or virtual) meeting. This conversation will typically allow both parties to agree on 1-3 final use-cases. Often this conversation will also yield a few “new” use-cases for consideration. Zapata will then take the information generated in this discussion and do some more homework to propose approaches to the use-cases of interest. A final discussion between our technical teams will then result in an agreement on one to create a proposal or Statement-of-Work that defines a proof of concept for your consideration. The summary is: we build solutions to real business problems, at scale, with cutting-edge techniques and quantum-inspired hardware. Additional time with Zapata Software and Science Experts is also available on request. Additionally, connectivity to third party hardware is included in your license. As part of an Enhanced Training package, Zapata experts can additionally consult to support purchasing time on third party devices and help develop a data-driven quantum purchasing strategy. You can request info on our training packages by emailing us at email@example.com.
Regarding algorithms: We have released basic VQE, QAOA, and QCBM implementations as examples of how to use Orquestra to deploy quantum algorithms as workflows. Moving forward we will be releasing new algorithms for chemistry, machine learning, optimization, finance, and other applications. Also, keep in mind that you can use Orquestra to deploy open-source algorithms or algorithms you have developed yourself. As for the cost model, we offer different packages depending on the need of the customer, including access to Orquestra, time with our Quantum Scientists, and more. Contact us at firstname.lastname@example.org for more information.
Yes, you can quickly switch to different data analysis tools without re-running the workflow. Because the results are returned in a JSON file, it’s not necessary to re-run the workflow to analyze it in a different way. We’ve open-sourced some of our tools for doing just that in our py-qe repo. See the README here for more details.
What’s great about Orquestra in that regard is that rather than being hardware-agnostic, Orquestra is hardware-smart! In other words, you can run your code at a high level of abstraction, just allowing Orquestra to do everything for you. Or, you can drill down and look at hardware-specific configurations, options, and specs if that’s what you need to do. Orquestra offers unparalleled flexibility in that area!
Yes, for example, check out our open-source Qiskit integration. You can specify a device name (and more!) when using the backend.
You can absolutely develop new quantum algorithms with Orquestra. Any code written in the popular open-source frameworks listed in the “Compose” section here can be brought in as an Orquestra Component. Because of that, algorithms utilizing PyQuil or Cirq for instance can also be used in Orquestra. Furthermore, Orquestra can accelerate the development of algorithms because you don’t have to build everything from scratch. Many algorithms share basic subroutines (consider for example the hybrid quantum-classical optimization loop in variational algorithms). These subroutines can be encoded as steps, allowing you to combine existing subroutines with new ones more easily in Orquestra to create, test and deploy new algorithms.
I hope this information has been helpful and please reach out to me here if you have any questions!
Sharing our understanding of the current state of AI and quantum.