Periodic Reporting for period 1 - PASQuanS2.1 (Programmable Atomic Large-scale Quantum Simulation 2 - SGA1)
Reporting period: 2023-04-01 to 2024-09-30
PASQuanS2.1 aims to pave the way towards the development of programmable atomic quantum simulators, such that in 2030 quantum simulators of up to 10000 quantum constituents are available; building an ecosystem for quantum simulations across Europe and serving as an interconnect between academia, start-ups and industrial end users. We aim to address important problems in fundamental and material science, as well as real-world problems of importance in industry. To this end, the consortium will build the next generation of programmable, large-scale atomic quantum simulators including the software stack with methods for verification, algorithm implementation, and user interfaces. It will research methods to validate the results obtained on these platforms, which sets the basis to demonstrate a practical advantage on them. At the same time, PASQuanS2.1 engages with end-users to develop new applications across science and industry for end-user applications and to expand industry connections to technology partners and startups.
PASQuansS2.1 will address four central challenges. First, the platforms for quantum simulation need to be technologically advanced. We aim to realize quantum simulators with at least 2000 individual quantum systems, with increasing calibration and a reduction of temperature and noise, and with full control over their interactions and couplings. This allows for more advanced programmable quantum simulations. Second, a software stack for control of these devices for broad applications needs to be developed. This connects at the highest level to end-user interfaces, and including systems engineering, verification and validation, including when the system is operating beyond regimes accessible to classical supercomputers. Relevant verification methods have been developed within PASQuanS (the predecessor project) and have to be integrated in the software stack and demonstrated for specific applications and one of the platforms. Third, industrial applications of quantum simulation need to be further developed so that quantum simulation extends beyond being an important scientific tool for physicists and develops its full potential towards maximum societal impact. This requires mapping real-life problems onto the dynamics of a simulator. From PASQuanS, there is a starting point with the implementation of optimization problems and the building of an end-user forum. The present challenge is to substantially expand this, identifying and developing implementations of new problem classes, while developing a sustainable ecosystem of end-users and open quantum simulation platforms. Forth, the consortium will facilitate the delivery of commercial quantum simulators, and in particular, cloud-based quantum simulators. This requires the consideration not only of the final platforms, but also the supply chain for components that build up quantum simulation platforms including system integration, allowing state-of-the-art commercial platforms to be made available online. It implies careful expansion of the IP portfolio and nurturing of start-ups in strategic areas.
PASQuansS2.1 will address four central challenges. First, the platforms for quantum simulation need to be technologically advanced. We aim to realize quantum simulators with at least 2000 individual quantum systems, with increasing calibration and a reduction of temperature and noise, and with full control over their interactions and couplings. This allows for more advanced programmable quantum simulations. Second, a software stack for control of these devices for broad applications needs to be developed. This connects at the highest level to end-user interfaces, and including systems engineering, verification and validation, including when the system is operating beyond regimes accessible to classical supercomputers. Relevant verification methods have been developed within PASQuanS (the predecessor project) and have to be integrated in the software stack and demonstrated for specific applications and one of the platforms. Third, industrial applications of quantum simulation need to be further developed so that quantum simulation extends beyond being an important scientific tool for physicists and develops its full potential towards maximum societal impact. This requires mapping real-life problems onto the dynamics of a simulator. From PASQuanS, there is a starting point with the implementation of optimization problems and the building of an end-user forum. The present challenge is to substantially expand this, identifying and developing implementations of new problem classes, while developing a sustainable ecosystem of end-users and open quantum simulation platforms. Forth, the consortium will facilitate the delivery of commercial quantum simulators, and in particular, cloud-based quantum simulators. This requires the consideration not only of the final platforms, but also the supply chain for components that build up quantum simulation platforms including system integration, allowing state-of-the-art commercial platforms to be made available online. It implies careful expansion of the IP portfolio and nurturing of start-ups in strategic areas.
Important progress has been made in the development of the quantum simulation platforms. For example, quantum simulators with thousands of neutral atoms prepared with fidelities up to 98% have been demonstrated. Better state preparation methods have been developed based on sorting of the atoms and improved potential control using spatial light modulators lead to extremely low temperatures. Readout capabilities have been enhanced by local basis rotations before readout. The reliability of the quantum simulation platforms improved in general, with a record of several weeks of autonomous operation. On the supply chain side, new ultra-low-noise laser sources with high output power have been developed and are available commercially.
The development of the software stack has been pushed forward. A reference implementation of a user-facing API has been developed, and programming environments are available within the consortium. For the verification of the platforms, digital twins based on different classical algorithms have been extended and improved. Towards the firmware side, we have developed optimal control methods to improve atom transport, gate implementations and state initialization.
We also worked on algorithm development, concentrating on variational methods, improved fermion-to-qubit mappings for quantum chemistry, and hardware-tailored solutions for diagonalization circuits. Limitations of certain popular algorithms and noise mitigation methods have been identified. For benchmarking, of quantum simulators, new sampling schemes from tensor network ansatz states have been developed.
We moved forward towards the demonstration of quantum advantage. To this end, we have performed quantum simulations of magnetic Hamiltonians in non-equilibrium situations, which are beyond the reach of classical simulation methods. We simulated fluctuating hydrodynamics in chaotic quantum systems for the accurate measurements of equilibrium transport coefficients using far-from-equilibrium dynamics. In Fermi-Hubbard systems, multipoint spin-spin and spin-charge correlations have been measured at different doping levels and at low temperatures. These results are obtained in a regime challenging for classical simulations and improve the understanding of the most popular model for strongly correlated electron systems.
Expansion of the end-user base and the development of new industry-academia collaborations was in the focus of our work. New partners from industry joined the consortium and the stakeholders board. To further intensify the collaboration, we initiated yearly meetings, and we are developing a common knowledge base around quantum simulators.
The development of the software stack has been pushed forward. A reference implementation of a user-facing API has been developed, and programming environments are available within the consortium. For the verification of the platforms, digital twins based on different classical algorithms have been extended and improved. Towards the firmware side, we have developed optimal control methods to improve atom transport, gate implementations and state initialization.
We also worked on algorithm development, concentrating on variational methods, improved fermion-to-qubit mappings for quantum chemistry, and hardware-tailored solutions for diagonalization circuits. Limitations of certain popular algorithms and noise mitigation methods have been identified. For benchmarking, of quantum simulators, new sampling schemes from tensor network ansatz states have been developed.
We moved forward towards the demonstration of quantum advantage. To this end, we have performed quantum simulations of magnetic Hamiltonians in non-equilibrium situations, which are beyond the reach of classical simulation methods. We simulated fluctuating hydrodynamics in chaotic quantum systems for the accurate measurements of equilibrium transport coefficients using far-from-equilibrium dynamics. In Fermi-Hubbard systems, multipoint spin-spin and spin-charge correlations have been measured at different doping levels and at low temperatures. These results are obtained in a regime challenging for classical simulations and improve the understanding of the most popular model for strongly correlated electron systems.
Expansion of the end-user base and the development of new industry-academia collaborations was in the focus of our work. New partners from industry joined the consortium and the stakeholders board. To further intensify the collaboration, we initiated yearly meetings, and we are developing a common knowledge base around quantum simulators.
Several experiments have been performed, advancing the state-of-the-art in the quantum simulation of strongly correlated many-body systems relevant to material science. This includes the simulation of non-equilibrium phenomena in quantum magnets and the study of non-equilibrium fluctuation-dissipation relations in strongly interacting itinerant systems. Important progress has also been made in simulations of strongly correlated Fermion systems in equilibrium. The observation of stripe formation and the measurement of correlations between different degrees of freedom up to 5th order provides new insight into the physics of these model systems.
Our advances beyond the state of the art are also visible in our IP protection activities. Up to today, three patents have been applied for by the consortium partners.
Our advances beyond the state of the art are also visible in our IP protection activities. Up to today, three patents have been applied for by the consortium partners.