Periodic Reporting for period 1 - NeQST (NExt level Quantum information processing for Science and Technology)
Okres sprawozdawczy: 2022-11-01 do 2024-04-30
Quantum computers and simulators exploit the laws of quantum mechanics to realize a novel computing paradigm. The hope is that superpositions and entanglement of quantum states allow for a much more efficient computation for specific but highly relevant applications.
Most contemporary quantum information processing platforms use two levels, ∣0⟩ and ∣1⟩, called qubits. The aim of NeQST is to elevate this to the next level, by developing the full stack of quantum information processing using d>2 levels, ∣0⟩,∣1⟩,…, ∣d−1⟩: qudits. This approach permits us to compactify quantum information and to tackle problems whose natural description is in terms of d-dimensional degrees of freedom, without encoding overheads.
NeQST tackles all levels of a quantum computer, (i) control of the underlying hardware, (ii) software control packages and simulators, (iii) algorithms for relevant use-cases, (iv) tools to certify the achieved results.
(i) In conventional trapped-ion quantum computers, quantum information is stored in two distinct states of an electron, denoted ∣0⟩ and ∣1⟩-the qubit. However, ions almost always have many more states that could be used to store information. The big opportunity-and challenge-of operating with qudits is the much richer structure to process the quantum information. Within NeQST, we develop new solutions to most efficiently harness the extended qudit spaces.
(ii) Recent accomplishments in the realization of quantum computers and simulators are leading to an exponentially increasing complexity, whose handling demands dedicated methods. The design automation for classical computers was a key to enable their ubiquity we see to today. We believe design automation has the potential to be a similar enabler for quantum computing. We want to push the boundaries on simulation, compilation, and verification as the pillars of design automation and apply them to qudits to fully exploit their potential.
(iii) NeQST will design algorithms for use cases where the additional freedom offered by qudits can become a game changer.
As academic use case, we tackle gauge theories, which describe fundamental particles of nature, such as electrons, positrons, quarks, or gluons. Despite their ubiquity, no efficient classical algorithm exists to compute their dynamical behavior. NeQST will design new algorithms for qudit quantum simulators.
As industrial use case, NeQST focuses on the challenges arising from the growing number of electric vehicles (EVs), which will substantially increase the load on the electricity grid and the variability of electricity production and consumption. Optimizing the distributed operations involved in the (dis-)charging schedules of large numbers of EVs requires novel algorithmic approaches. It is still an open question whether and at which scale quantum computers may offer a speedup in such a context. Yet, it is important to identify and enhance the possibilities quantum computers may offer.
(iv) Present quantum computers are noisy, making it particularly important to certify their quantum performance. In the context of the vastly increased complexity of qudits, this requires the development of new approaches and experimentally feasible tools. We develop techniques to certify that (a) the qudit quantum hardware works as intended, and (b) that it uses intrinsic qudit quantum phenomena.
Most contemporary quantum information processing platforms use two levels, ∣0⟩ and ∣1⟩, called qubits. The aim of NeQST is to elevate this to the next level, by developing the full stack of quantum information processing using d>2 levels, ∣0⟩,∣1⟩,…, ∣d−1⟩: qudits. This approach permits us to compactify quantum information and to tackle problems whose natural description is in terms of d-dimensional degrees of freedom, without encoding overheads.
NeQST tackles all levels of a quantum computer, (i) control of the underlying hardware, (ii) software control packages and simulators, (iii) algorithms for relevant use-cases, (iv) tools to certify the achieved results.
(i) In conventional trapped-ion quantum computers, quantum information is stored in two distinct states of an electron, denoted ∣0⟩ and ∣1⟩-the qubit. However, ions almost always have many more states that could be used to store information. The big opportunity-and challenge-of operating with qudits is the much richer structure to process the quantum information. Within NeQST, we develop new solutions to most efficiently harness the extended qudit spaces.
(ii) Recent accomplishments in the realization of quantum computers and simulators are leading to an exponentially increasing complexity, whose handling demands dedicated methods. The design automation for classical computers was a key to enable their ubiquity we see to today. We believe design automation has the potential to be a similar enabler for quantum computing. We want to push the boundaries on simulation, compilation, and verification as the pillars of design automation and apply them to qudits to fully exploit their potential.
(iii) NeQST will design algorithms for use cases where the additional freedom offered by qudits can become a game changer.
As academic use case, we tackle gauge theories, which describe fundamental particles of nature, such as electrons, positrons, quarks, or gluons. Despite their ubiquity, no efficient classical algorithm exists to compute their dynamical behavior. NeQST will design new algorithms for qudit quantum simulators.
As industrial use case, NeQST focuses on the challenges arising from the growing number of electric vehicles (EVs), which will substantially increase the load on the electricity grid and the variability of electricity production and consumption. Optimizing the distributed operations involved in the (dis-)charging schedules of large numbers of EVs requires novel algorithmic approaches. It is still an open question whether and at which scale quantum computers may offer a speedup in such a context. Yet, it is important to identify and enhance the possibilities quantum computers may offer.
(iv) Present quantum computers are noisy, making it particularly important to certify their quantum performance. In the context of the vastly increased complexity of qudits, this requires the development of new approaches and experimentally feasible tools. We develop techniques to certify that (a) the qudit quantum hardware works as intended, and (b) that it uses intrinsic qudit quantum phenomena.
The NeQST team has made progress on all of the levels mentioned above.
Methods for efficiently characterizing qudits and qudit gate sets were developed. A particular focus was on hardware and software control to stabilize the more complex Hilbert space against new kinds of errors. A qudit programming language was developed that now forms the basis for algorithm development within the consortium. A classical simulator platform was developed to enable the validation against experimental noise models, and a software package was published as open source.
Blueprints were designed for using the NeQST qudit architecture for quantum simulating gauge theories, and a small version of quantum electrodynamics in two spatial dimensions was experimentally simulated. These advances confirm the suitability of the NeQST hardware for this use case and they lay the foundation for proceeding to more complex theories. Moreover, we have designed protocols for making such quantum simulations resilient against unavoidable device errors.
The most relevant aspects of realistic electric vehicle (EV) charging optimization problems were analyzed and several problem variants were formulated in terms of qudits. We have developed methods for addressing the many inequality constraints and multiple objectives that typically characterize realistic problems. These methods and analyses form the basis for flexible quantum optimization algorithms.
Finally, we have studied different protocols for evaluating the performance of qudit platforms. Ways have been developed to benchmark the performance of quantum devices in quantum optimization by certifying the amount of entanglement generated. Methods for certification of the dimension of a physical system were implemented experimentally. A quantum-inspired variational algorithm to study integer optimization problems was derived, showing competitive or even better results than other state-of-the-art methods.
Methods for efficiently characterizing qudits and qudit gate sets were developed. A particular focus was on hardware and software control to stabilize the more complex Hilbert space against new kinds of errors. A qudit programming language was developed that now forms the basis for algorithm development within the consortium. A classical simulator platform was developed to enable the validation against experimental noise models, and a software package was published as open source.
Blueprints were designed for using the NeQST qudit architecture for quantum simulating gauge theories, and a small version of quantum electrodynamics in two spatial dimensions was experimentally simulated. These advances confirm the suitability of the NeQST hardware for this use case and they lay the foundation for proceeding to more complex theories. Moreover, we have designed protocols for making such quantum simulations resilient against unavoidable device errors.
The most relevant aspects of realistic electric vehicle (EV) charging optimization problems were analyzed and several problem variants were formulated in terms of qudits. We have developed methods for addressing the many inequality constraints and multiple objectives that typically characterize realistic problems. These methods and analyses form the basis for flexible quantum optimization algorithms.
Finally, we have studied different protocols for evaluating the performance of qudit platforms. Ways have been developed to benchmark the performance of quantum devices in quantum optimization by certifying the amount of entanglement generated. Methods for certification of the dimension of a physical system were implemented experimentally. A quantum-inspired variational algorithm to study integer optimization problems was derived, showing competitive or even better results than other state-of-the-art methods.
The NeQST project has significantly progressed beyond the state of the art in all addressed aspects.
Quantum control improvements have permitted the quantum processor to tackle larger problem sizes with improved reliability of components, both crucial requirements towards solving relevant application problems. Optimal single qudit design solutions reduce the average number of single-qudit operations by a factor of 2. A design-automation software suite was released to the broader community as open access. It provides a standardized, easy-to-use, and flexible framework for the development of qudit algorithms and their simulation, which is expected to broadly accelerate the end-user uptake.
The developed algorithms for gauge theory quantum simulation permit to address significantly more complex problems, e.g. multiple flavors and dimensions larger 1. The developed strategies for error mitigation increase the reliability of the quantum hardware, and they have revealed a previously unknown transition that sharply divides an error-protected and an unprotected regime.
The designed formulations for the EV-charging problem permit to tackle this important use case in quantum hardware. In the long run, by optimizing a flexible reservoir of EV-batteries, this may improve the use of variable green energy, reduce the stress on the electricity grid, foster the transition to a local grid, and increase the participation of the general public in the electricity market. The methods designed to tackle a variety of realistic constraints and multiple objectives in quantum optimization will also be important for applications beyond the immediate context of the project.
Quantum control improvements have permitted the quantum processor to tackle larger problem sizes with improved reliability of components, both crucial requirements towards solving relevant application problems. Optimal single qudit design solutions reduce the average number of single-qudit operations by a factor of 2. A design-automation software suite was released to the broader community as open access. It provides a standardized, easy-to-use, and flexible framework for the development of qudit algorithms and their simulation, which is expected to broadly accelerate the end-user uptake.
The developed algorithms for gauge theory quantum simulation permit to address significantly more complex problems, e.g. multiple flavors and dimensions larger 1. The developed strategies for error mitigation increase the reliability of the quantum hardware, and they have revealed a previously unknown transition that sharply divides an error-protected and an unprotected regime.
The designed formulations for the EV-charging problem permit to tackle this important use case in quantum hardware. In the long run, by optimizing a flexible reservoir of EV-batteries, this may improve the use of variable green energy, reduce the stress on the electricity grid, foster the transition to a local grid, and increase the participation of the general public in the electricity market. The methods designed to tackle a variety of realistic constraints and multiple objectives in quantum optimization will also be important for applications beyond the immediate context of the project.