Now renamed to IBM Quantum, it has over 360,000 registered users, with an average of 2.2B circuit shots executed per day in 2021 and typically 25 quantum processors available online at any one time. of Bristol in 2013 ), but it was the launch of the IBM Quantum Experience in 2016 that really succeeded in driving a step change in engagement. It was others who first put quantum processors on the cloud (Jeremy O’Brien at the Univ. IBM has been the clear early victor in quantum software’s ‘initial awareness’ phase. To understand the current state of play, investors have to appreciate the challenges and opportunities faced by the early players, and the different longer-term commercial strategies towards which they are evolving. For early adopters and would-be quantum developers, the most important consideration may not be immediately available software product features, but picking the right partners. However, Fact Based Insight’s reviews of progress on quantum hardware and quantum algorithms indicate that this is still likely to be a long game. The development of the quantum software sector is proceeding rapidly. They represent the status of which Fact Based Insight is aware and we caution readers to take them only as a starting point for their own further evaluations. Note: by design the diagrams supporting this briefing represent many different approaches against one simplified view of the stack. Simulator – Conventional simulators are a key additional element of the software stack, not just because of the limited performance of current quantum processors, but also to support ongoing programme development and debugging. For a discussion focussed on the quantum layer see Quantum Hardware Outlook 2022. Advanced protocols are required to optimise gate operations and to suppress crosstalk. Control – Low level operations driven by analogue pulses (typically microwave or laser based). In the end we can expect kernels optimised for the efficient implementation of quantum error correction, and that co-ordinate access to specialised resources such as magic state factories and QRAM. Includes quantum gates, measurements and tightly coupled classical logic. Architecture – Runtime environment co-ordinating compute operations. Various providers offer their own way of describing and executing the required quantum circuits. Framework – Most early players emphasise circuit-model quantum computation (though there are important variations on this approach). For a detailed discussion see Quantum Algorithms Outlook 2022. Algorithms – Unique quantum approaches to solve various classes of problem. Many players have emphasised early research and user community engagement to develop proof-of-concept and trial applications. For the most part these remain a work in progress. Fact Based Insight’s simplified view of the quantum stack is inspired by the activities of early players and lessons from the evolution of the modern HPC stack.Īpplications – End-use business-oriented applications. There is no layer in the quantum stack that doesn’t deserve our close attention. At the lowest level qubits require analogue control, itself a rich seam of challenge and opportunity. Error correction is a key part of the quantum story and also needs to find its place. Finding the right framework to cast such maths into program form is also key, but the circuits of gate-model quantum computing and quantum annealing solvers aren’t the only games in town. Innovative quantum algorithms are equally important, so making these easier to develop and deploy for real business applications will be a central concern. The quantum revolution relies on much more than just building a new form of computing hardware.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |