Born at Nobel Symposium 210 · Quantum Software Lab
Primitive–Platform Matchmaking
Tell us about your machine. We map the algorithmic primitives onto it and send you a matchmaking report: candidate primitive constructions for your machine, the assumptions to check, and experiments that would show them.
This is an open, academic pilot run by the Quantum Software Lab (University of Edinburgh) with the research community. It is non-commercial: nothing you share is used for any commercial purpose. We ask only for knowledge that is already in the public domain, such as published papers, talks and public specifications. Any level of participation is welcome, from a single paragraph to a conversation with your team.
Your academic research system. We ask about your academic, white-box system: the most advanced one in your lab today. We do not ask about any commercial platform, product or company you may be connected to.
Not a benchmark. We match primitives to your platform as it is today and suggest candidate constructions for it. It is a snapshot of a current academic platform, not a score, a ranking or a verdict on your design.
Public results are about primitives, not platforms. What each primitive needs from hardware is open on the Atlas, so any group can design towards it. Nothing about your platform is published without your approval.
Use it both ways. If you design for error correction and scale independently of applications, the open requirements can help you adjust your design; tell us when it changes and we update your match. There is no exclusivity and no long-term commitment: update, switch or leave whenever you wish.
Born at Nobel Symposium 210
This pilot was conceived for, and launched at, Nobel Symposium 210 in Stellenbosch, in October 2026. The symposium brought the people who design quantum algorithms and the people who build quantum machines into the same room for a week. The question raised there, which primitive each machine runs natively, is the question this pilot tries to answer. We are grateful to the symposium for making that conversation possible, and we hope the matchmaking keeps it going.
Why
Algorithms are built from a small set of primitives, each defined by a contract: what it needs and what it guarantees. Machines differ in what they make cheap. Two devices on the same physics can favour different primitives because they scale differently: shuttling, photonic links, microwave control. Today the match between primitives and hardware is made case by case. We want to make it systematic, on correct data, with the people who build the machines.
A note on words: in the Atlas, an "architectural primitive" is a family of circuits, such as a quantum convolutional network. Your machine gets its own hardware card.
How it works
Describe your hardware and its cheap and costly operations, in your own words. Point us to papers, or a person.
Read your input with the Atlas cards and the literature, and match primitives and their variants to your machine.
Receive your report first, in early 2027. You correct it, and you decide what, if anything, becomes public.
What we ask: eight parts
- Your configuration. Your academic research system as it is today (not a commercial platform): group, a named configuration, a public reference, and where the report goes.
- Physical structure and interactions. Carriers, and the interactions available today.
- Preparing, using and keeping states and data. Copies, memory, inverse and coherently controlled operations, data access.
- Readout and control during a run. Measurement, reset, feedback, detection.
- What enters, leaves and connects. Classical and quantum input and output, modules, and what crosses a connection: links that matter for verification and for joining communication and computing.
- What to explore today, and your design intent, kept separate from today's snapshot.
- What you would most like a theorist to tell you.
- Your consent.
Parts 2 to 5 are your Lego pieces: tick what your machine offers today. For up to two things that matter most in each part, tell us whether it is native, routine or costly on your machine today, and how your machine does it, in a few words or with a paper link. No numbers, no weights. You do not need to choose primitives: the matchmaking is our job. Please share only what is already public, and nothing under a non-disclosure agreement or export control.
What you get
- A matchmaking report for your machine in early 2027: candidate primitive constructions, the assumptions still to check, and experiments that would show them.
- A hardware card for your machine in the Atlas, credited to your group, once you approve it.
- Searches by our algorithm engine, where feasible, aimed at your machine's native operations.
- Answers to your theory questions on algorithms, primitives, complexity or verification, as part of the matchmaking.
- An invitation to co-author the joint white paper, for groups who contribute substantially.
How we work together
- Open and non-commercial. An academic pilot of the Quantum Software Lab. Nothing you share is used for commercial purposes.
- Your academic system only. We ask about your academic, white-box research system, not about any commercial platform, product or company you may be connected to.
- Public knowledge only. We ask only for what is already in the public domain. Please do not send anything under a non-disclosure agreement or export control.
- Findings go to you first. Each group receives its own report before anything based on its answers is shared.
- Nothing based on your answers is announced without your approval. No statement drawn from your answers is published, presented or posted without your written approval.
- You can withdraw. Write to the editor at any time and we delete your answers within 30 days. This does not apply to material you have already approved for publication.
- You keep your data and may use the findings freely in your own papers, proposals and talks.
- Not a benchmark. We match primitives to your current platform. We will not publish a ranking, a score or a comparison of platforms, and public results describe what primitives need, not how platforms perform.
- No exclusivity. No long-term commitment and no exclusive partnership: update your answers, work with others, or leave whenever you wish.
- Credit. Contributors are named on the Atlas and on their hardware card if they wish. The criteria for co-authorship of the joint white paper are sent to all contributors before any paper is drafted; other contributions are acknowledged.
- Use of AI tools. AI agents working for the editor help us read published papers, read your answers, build the cards and draft reports. Since you share only public knowledge, your technical answers may be used this way. Your contact details are used only to write to you, and every result is checked by a scientist.
- The Atlas stays open. It is public, under a Creative Commons Attribution licence (CC BY 4.0), and contributors help shape its next versions.
Privacy notice
- Who holds your data: Elham Kashefi, editor of the Atlas, Quantum Software Lab, University of Edinburgh.
- What and why: your name, email, group and answers, used only to prepare your matchmaking report, your hardware card and contributor credit.
- Legal basis: your consent, given in the form. You can withdraw it at any time.
- Where it is stored: in Microsoft Forms under the University of Edinburgh's Microsoft 365 account, and in the editor's research files. No other service is used. Your answers are read by the editor and by AI agents working for the editor, as described in the charter above.
- How long: for the duration of the pilot, and at most three years after you answer; earlier if you ask.
- Your rights: to see, correct or delete your data, and to withdraw consent, by writing to the editor. You may also complain to the UK Information Commissioner's Office or the data protection authority of your country.
- A team contact: please give a colleague's details only with their agreement.
Contact: Elham Kashefi, ekashefi@exseed.ed.ac.uk. Please do not send confidential material by email.