Past seminars

Find details and recordings of our previous seminars.

Past Seminars

Find the abstracts for our past seminars below. You can also watch videos of all our talks on our YouTube Channel.

12 July 2022 | 16:00 UTC | Matthias Werner | Qilimanjaro

A graph-theoretic analysis on first order quantum phase transitions for adiabatic quantum computing

In the context of adiabatic quantum computation (AQC), it has been argued that first order quantum phase transitions (QPTs) due to localisation phenomena will always cause AQC to fail by exponentially decreasing the minimal spectral gap of the Hamiltonian along the annealing path. The vanishing spectral gap occurs due to localisation of the ground state in a local minimum, requiring the system to tunnel into the global minimum at a later stage of the annealing. However this notion has been subject to some debate in the community, since more recent findings suggest the existence of methods to avoid this by carefully designing the involved Hamiltonians. It remains a challenge to formulate a comprehensive theory on the effect of the various parameters and the conditions under which QPTs make the AQC algorithm fail. In this work we investigate the conditions under which localisation causes first order QPTs using spectral graph theory, examine both analytically and numerically the role of the connectivity of the driver Hamiltonian in the mitigation of such effects in different AQC algorithms and derive bounds on the location of the minimal spectral gap along the anneal path. Additionally, we show that in the limiting case of a fully connected driver Hamiltonian as used in adiabatic Grover search, first order QPTs due to localisation are avoided entirely. Our analysis augments the tool box to design the driver such that first order QPTs are avoided and the runtime of AQC algorithms is improved.

26 July 2022 | 00:01 UTC | Arnab Banerjee | Purdue University

Opportunities in magnetic ground state discovery using annealers and beyond

The knowledge of the ground states, the phase diagram and the critical properties of quantum magnets, especially those which can yield interesting and unconventional phases, is essential to enable future quantum spintronics technologies. Dynamic susceptibility experiments, such as neutron scattering, heat capacity, and magnetic susceptibility, performed on materials often show a variety of ground and excited states, which are often difficult to understand. Ising quantum annealers, quantum-inspired parallel tempering machines and certain universal ion-trap-based devices performing analog simulations have started to reach the level of maturity required to be useful tools for solving these problems without having to retort to conventional monte-carlo routines. In this talk, we describe the solution to the frustrated Ising Shastry-Sutherland Hamiltonian using three different backends - D-Wave, Fujitsu and Honeywell. The Hamiltonian is arguably expressed in several rare-earth magnets, such as Erbium tetraborides which provides a physics motivation to solving this problem. We showcase that each backend has its opportunities and drawbacks which if used appropriately, can be a powerful tool to access a variety of phases, phase transitions, as well as their (quantum) critical behaviour.

9 August 2022 | 08:00 UTC | Hidetoshi Nishimori | Tokyo Institute of Technology

Quantum optimization of a continuous- variable function with rugged energy landscape

Quantum annealing is applied to an optimization problem with a continuous degree of freedom. The energy landscape has a number of local minima, and simulated annealing converges logarithmically slowly. We show by extensive numerical analyses that quantum annealing has power law convergence, thus an exponential improvement over simulated annealing. We also display movies illustrating how quantum tunneling steers the systems toward global minimum

23rd August 2022 | 16:00 UTC | Glen Mbeng | University of Innsbruck

Designing quantum annealing schedules with the counterdiabatic rotated ansatz

Approximate counterdiabatic (CD) protocols are a powerful tool to enhance quantum adiabatic processes that allow for manipulating quantum systems on short time scales. However, implementing CD protocols entails the introduction of additional control fields in the Hamiltonian, often associated with highly non-local multi-body interactions. We will introduce a novel variational rotated ansatz (RA) to generate experimentally accessible approximate CD protocols. We will show that the RA provides a framework to improve quantum annealing schedules without having access to spectral information. We will present numerical benchmarks for different annealing architectures, demonstrating the benefits of RA protocols and exemplifying their capability to enhance the algorithms’ success probability.

7th September 2022 | 00:01 UTC | Xi Dai | University of Waterloo

Dissipative Landau Zener Tunneling: crossover from weak

The Landau-Zener problem for a two-level system is a suitable toy problem for studying quantum tunnelling in an annealer. Coupling to the environment can influence the tunnelling probability and theoretical understanding is only available for specific coupling limits or noise models. We present experimental results on Landau-Zener measurements on a capacitively-shunted flux qubit. The result shows crossover from weak to strong coupling to the environment. Our results in the weak and strong coupling limits are consistent with previous theoretical and numerical studies and our result in the intermediate regime is novel. Understanding the intermediate regime can guide novel designs and operations of annealers, utilizing enhanced control capabilities and/or engineered environment, which could improve the success probabilities of quantum annealing.

20th September 2022 | 08:00 UTC | Dr Yuichi Igarashi | NEC

Toward the realization of a quantum annealing machine based on Josephson parametric oscillators

Josephson parametric amplifier (JPA) is commonly used to read out a weak signal of a superconducting qubit. When a pump power of the JPA exceeds a threshold, it starts to oscillate even without an input signal. In that case the device is called a Josephson parametric oscillator (JPO). The oscillation states of a JPO composed of two different states with the same amplitude but opposite phase (0 pi or 1 pi) can be used as two basis states of a qubit. We study basic physical properties of a JPO toward the realization of a quantum annealing machine based on JPOs. We also develop three-dimensional packaging technologies, which are necessary for the large-scale integration in the future.

27th September 2022 | 16:00 UTC | Adolfo del Campo | University of Luxembourg

Universal dynamics of open quantum phase transitions in a quantum annealer

The number of topological defects created in a system driven through a quantum phase transition exhibits a power-law scaling with the driving time. This universal scaling law is the key prediction of the Kibble-Zurek mechanism (KZM), and testing it using a hardware-based quantum simulator is a coveted goal of quantum information science. Here we provide such a test using quantum annealing. Specifically, we report on extensive experimental tests of topological defect formation via the one-dimensional transverse-field Ising model on two different D-Wave quantum annealing devices. We find that the quantum simulator results can indeed be explained by the KZM for open-system quantum dynamics with phase-flip errors, with certain quantitative deviations from the theory likely caused by factors such as random control errors and transient effects. In addition, we probe physics beyond the KZM by identifying signatures of universality in the distribution and cumulants of the number of kinks and their decay, and again find agreement with the quantum simulator results. This implies that the theoretical predictions of the generalized KZM theory, which assumes isolation from the environment, applies beyond its original scope to an open system. We support this result by extensive numerical computations. To check whether an alternative, classical interpretation of these results is possible, we used the spin-vector Monte Carlo model as well as the spin-vector Langevin model, both candidate classical descriptions of the D-Wave device. Our work provides an experimental test of quantum critical dynamics in an open quantum system, and paves the way to new directions in quantum simulation experiments.

5th October 2022 | 00:01 UTC | Eliot Kapit | Colorado School of Mines

Noise tolerant quantum speedups in quantum annealing without fine tuning

Quantum annealing (QA) is a promising method to solve hard optimization problems with quantum hardware, without the need for fault tolerance and topological error correction. However, despite great effort it has thus far failed to achieve broadly applicable quantum advantage in practical problems. We identify four key issues as the likely reason for this, two of which are engineering challenges and two of which are deeper physics problems. We propose a novel modification, called RFQA, which we argue will solve or at least mitigate the core physics issues in ordinary QA. This modification applies low-frequency oscillating terms independently to every qubit in the system, which results in an exponential proliferation of weak resonances that accelerates the first order phase transitions that bottleneck QA. This novel quantum speedup mechanism allows for faster thermalization in glassy problems; we present a mix of analytical and numerical results demonstrating this. Implemented at scale, RFQA would thus be an extremely promising route to achieving near-term quantum advantage in practical problems.

11th October 2022 | 08:00 UTC | Nobuyuki Yoshikawa | Yokohama National University

Energy-efficient superconductor logic for interfacing with quantum annealing systems

The recent rapid increase in the scale of superconducting quantum computing systems greatly increases the demand for qubit control by digital circuits operating at qubit temperatures. In this paper, superconducting digital circuits, such as single-flux quantum and adiabatic quantum flux parametron circuits are described, that are promising candidates for this purpose. After estimating their energy consumption and speed, a conceptual overview of the superconducting electronics for controlling a multiple-qubit system is provided, as well as some of its component circuits.

18th October 2022 | 16:00 UTC | Filip Wudarski | USRA

Application specific annealers: A case study for Fermi-Hubbard annealers.

Quantum annealers have been broadly studied in the context of combinatorial optimization problems onto which you can map a host of interesting problems ranging from traffic optimization to machine learning. However, these annealers are not universal, and fail to exploit the full power of quantum computing. Therefore, in this talk, we will present an alternative look at annealing that is tailored to solve a different class of problems, namely Fermi-Hubbard models (FHMs). FHMs are proxy models for material simulations that allow us to study phenomena such as superconductivity or insulating phases. The problem Hamiltonians (given in second quantization) first need to be mapped to the qubit structure, which we achieve by employing a low-weight encoding requiring at most 3-local interactions. We design an annealing protocol that is investigated numerically and provides promising results for spinless and spinful systems. Additionally, we will present experimental perspectives for the proposed design and compare the number of resources required to solve the same problems with a gate based quantum computers and Fermi-Hubbard Annealers.

1st November 2022 | 09:00 UTC | Siya Bao | Waseda University

A quantum computing-based optimization method for multi-day travel recommendation

The multi-day travel planning assists users with realistic travel itineraries by searching for the optimal travel routes through a set of candidate hotels and point-of-interests (POIs). The multi-day travel planning problem (MTPP) can be solved as an optimization problem. Although conventional methods using von Neumann computers obtain good approximate solutions to the optimization problems, large computation costs are required to solve large-scale or complex problems due to the combinatorial explosion. On the other hand, Ising machines or quantum annealing machines are non-von Neumann computers, and those machines are developed to deal with complex optimization problems. In this paper, we propose an Ising-machine-based method for the MTPP. Practical factors of the MTPP include the POI satisfaction, travel expenses, and time limits. Those factors are mapped onto quadratic unconstrained binary optimization (QUBO) forms. We evaluate the proposed method using two real-world datasets including Sapporo and Tokyo, Japan. Experimental results show that the MTPP can be effectively solved using Ising machines compared with the conventional methods in terms of the solution quality and the execution time. To the best of our knowledge, this study is the first solution of the MTPP using Ising machines.

16th November 2022 | 01:00 UTC | Lucas Brady | NASA

Simultaneous Stoquasticity

Stoquastic Hamiltonians play a role in the computational complexity of the local Hamiltonian problem as well as the study of classical simulability. In particular, stoquastic Hamiltonians can be straightforwardly simulated using Monte Carlo techniques. We address the question of whether two or more Hamiltonians may be made simultaneously stoquastic via a unitary transformation. This question has important implications for the complexity of simulating quantum annealing where quantum advantage is related to the stoquasticity of the Hamiltonians involved in the anneal. We find that for almost all problems no such unitary exists and show that the problem of determining the existence of such a unitary is equivalent to identifying if there is a solution to a system of polynomial (in)equalities in the matrix elements of the initial and transformed Hamiltonians. Solving such a system of equations is NP-hard. We highlight a geometric understanding of this problem in terms of a collection of generalized Bloch vectors. [Phys. Rev. A 105, 062601]

22nd November 2022 | 09:00 UTC | Yuichiro Matsuzaki | AIST

Obtaining Ground States of the XXZ Model Using the Quantum Annealing with Inductively Coupled Superconducting Flux Qubits

Obtaining ground states of Hamiltonians is important in the condensed matter physics and the quantum chemistry. The interaction Hamiltonians typically contain not only diagonal but also off-diagonal elements. Although quantum annealing provides a way to obtain a ground state of a Hamiltonian, we can only use the Hamiltonian with Ising interaction by using currently available commercial quantum annealing devices. In this work, we propose a quantum annealing for the XXZ model, which contains both Ising interaction and energy-exchange interaction, by using inductively coupled superconducting flux qubits. The key idea is to use a recently proposed spin-lock quantum annealing where the qubits are driven by microwave fields. As long as the rotating wave approximation is valid, the inductive coupling between the superconducting flux qubits produces the desired Hamiltonian in the rotating frame, and we can use such an interaction for the quantum annealing while the microwave fields driving play a role of the transverse fields. To quantify the performance of our scheme, we implement numerical simulations, and show that we can obtain ground states of the two-dimensional Heisenberg model with a high fidelity.

29th November 2022 | 17:00 UTC | Emanuele Dalla Torre | Bar-Ilan University

Navigating the limits of adiabatic quantum computers

Adiabatic quantum algorithms solve computational problems by slowly evolving a trivial state to the desired solution. On an ideal quantum computer, the solution quality improves monotonically with increasing circuit depth. By contrast, increasing the depth in current noisy computers introduces more noise and eventually deteriorates any computational advantage. What is the optimal circuit depth that provides the best solution? Here, we address this question by investigating an adiabatic circuit that interpolates between the paramagnetic and ferromagnetic ground states of the one-dimensional quantum Ising model. We characterize the quality of the final output by the density of defects d, as a function of the circuit depth N and noise strength σ. We find that d is well-described by the simple form dideal+dnoise, where the ideal case dideal∼N−1/2 is controlled by the Kibble-Zurek mechanism, and the noise contribution scales as dnoise∼Nσ2. It follows that the optimal number of steps minimizing the number of defects goes as ∼σ−4/3. We implement this algorithm on a noisy superconducting quantum processor and find that the dependence of the density of defects on the circuit depth follows the predicted non-monotonous behavior and agrees well with noisy simulations. Our work allows one to efficiently benchmark quantum devices and extract their effective noise strength σ.

7th December 2022 | 01:00 UTC | David Ferguson | NGC

Requirements for Non-Stoquasticity in Superconducting Circuits

Stoquasticity Allows the Ground State Quantum Systems To Be “Efficiently” Simulated by Classical Computers. Recent Couplers Have Generated Small +XX and YY Couplings But Are Fundamentally Stoquastic. In contrast Aharonov-Casher Effect Allows for Strong Non-Stoquastic Effects in Flux Qubits. This presentation reviews previous demonstrations of annealing qubits capable of strong non-stoquastic effects and discusses possible future demonstrations.

13th December 2022 | 09:00 UTC | Dr Tadashi Kadowaki | DENSO

Greedy parameter optimization for diabatic quantum annealing

A shorter processing time is desirable for quantum computation to minimize the effects of noise. We propose a simple procedure to variationally determine a set of parameters in the transverse-field Ising model for quantum annealing appended with a field along the y-axis. The method consists of greedy optimization of the signs of coefficients of the y-field term based on the outputs of short annealing processes. We test the idea in the ferromagnetic system with all-to-all couplings and spin-glass problems, and find that the method outperforms the traditional form of quantum annealing and simulated annealing in terms of the success probability and the time to solution, in particular in the case of shorter annealing times, achieving the goal of improved performance while avoiding noise. The non-stoquastic σy term can be eliminated by a rotation in the spin space, resulting in a non-trivial diabatic control of the coefficients in the stoquastic transverse-field Ising model, which may be feasible for experimental realization.

24th January 2023 | 09:00 UTC | Dr.Koji Mizumatsu | Fixstars

The Fixstars Amplify: Versatile programming platform for quantum annealing machines and Ising machines

At present, many quantum annealing machines and Ising machines have been proposed, and several machines have also been put into practical use. However, the software development environment and operation methods of each machine are very different, so it is not easy for researchers and engineers to use them. The Fixstars Amplify has been developed as a cloud platform to facilitate the development and execution of algorithms for solving combinatorial optimization problems of commercially available quantum annealing machines, Ising machines, mathematical optimization solvers, and gated quantum computers. An overview of its features will be introduced in this talk.

31st January 2023 | 17:00 UTC | Maddie Cain | Harvard University

Quantum Optimization of Maximum Independent Set with Rydberg Atom Arrays

Realizing quantum speedup for practically relevant, computationally hard problems is a central challenge in quantum information science. In this talk, I will present experimental investigations of quantum algorithms for solving the Maximum Independent Set problem using Rydberg atom arrays with up to 289 qubits in two spatial dimensions. I will outline how we use a hardware-efficient encoding associated with Rydberg blockade, realize closed-loop optimization to test several variational algorithms, and apply them to systematically explore a class of graphs with programmable connectivity. Next, I will discuss the results of benchmarking the quantum algorithm's performance against classical simulated annealing and explain graph properties that control the problem hardness. Finally, I will explain our observations of a superlinear quantum speedup on the hardest graphs in finding exact solutions in the deep circuit regime and analyze its origins.

8th February 2023 | 01:00 UTC | Luis Pedro Garcia-Pintos | Los Alamos National Laboratory

Lower Bounds on Quantum Annealing Times

The adiabatic theorem provides sufficient conditions for the time needed to prepare a target ground state. While it is possible to prepare a target state much faster with more general quantum annealing protocols, rigorous results beyond the adiabatic regime are rare. Here, I will provide such a result, showing lower bounds on the time needed to successfully perform quantum annealing. The bounds are asymptotically saturated by three toy models where fast annealing schedules are known: the Roland and Cerf unstructured search model, the Hamming spike problem, and the ferromagnetic p-spin model. The results also show that rapid annealing requires coherent superpositions of energy eigenstates, singling out quantum coherence as a computational resource.

14th February 2023 | 09:00 UTC | Akiyoshi Tomonaga | Riken

Architecture and their implementation for superconducting quantum annealer and NISQ circuit

The development of quantum computers has become an extremely important topic in recent years due to society's demand for accelerated growth in the amount of information. Our group has developed and implemented architectures for full-coupled quantum annealers and gate-based quantum computation circuits that can be implemented in a planar manner. We report on our latest results and efforts on these topics , including experiments on packaging and coherence enhancement.

21st February 2023 | 17:00 UTC | Michael Hanks | Imperial College London

Divide-and-conquer embedding for QUBO quantum annealing

Quantum annealing promises to be an effective heuristic for complex NP-hard problems. However, clear demonstrations of quantum advantage are wanting, primarily constrained by the difficulty of embedding the problem into the quantum hardware. While divide-and-conquer strategies have been applied in the more general context of quantum approximate optimization, they have been limited in QUBO-restricted experimental quantum annealing. Here, we exploit the sample properties obtained from annealing devices, deliberately worsening typical measures of embedding quality to improve the partial solutions we obtain for each subproblem. We apply our approach first to the highly irregular graph of an integer factorisation problem and, passing this initial test, move on to consider more regular geometrically frustrated systems. Our results show that a problem-focused approach to embedding can improve performance by orders of magnitude.

14th March 2023 | 17:00 UTC | Marcel Seelbach | University of Siegen

Quantum annealing with learnt couplings.

Modern quantum annealers can find high-quality solutions to combinatorial optimisation problems given as quadratic unconstrained binary optimisation (QUBO) problems. Unfortunately, obtaining suitable QUBO forms in computer vision remains challenging and currently requires problem-specific analytical derivations. Moreover, such explicit formulations impose tangible constraints on solution encodings. In stark contrast to prior work, this paper proposes to learn QUBO forms from data through gradient backpropagation instead of deriving them. As a result, the solution encodings can be chosen flexibly and compactly. We demonstrate the advantages of learnt QUBOs on the diverse problem types of graph matching, 2D point cloud alignment and 3D rotation estimation.

22nd March 2023 | 01:00 UTC | Tameem Albash | The University of New Mexico

Parallel Tempering for Ground State Approximation using Artificial Neural Networks

A large body of work has demonstrated that parameterized artificial neural networks (ANNs) can efficiently describe ground states of numerous interesting quantum many-body Hamiltonians. However, the standard variational algorithms used to update or train the ANN parameters can get trapped in local minima, especially for frustrated systems and even if the representation is sufficiently expressive. We propose a parallel tempering method that facilitates escape from such local minima. This methods involves training multiple ANNs independently, with each simulation governed by a different Hamiltonian, and it incorporates an update step into the training that allows for the exchange of neighboring ANN configurations. We study instances from two classes of Hamiltonians to demonstrate the utility of our approach using Restricted Boltzmann Machines as our parameterized ANN. The first instance is based on a permutation-invariant Hamiltonian whose landscape stymies the standard training algorithm by drawing it increasingly to a false local minimum. The second instance is four hydrogen atoms arranged in a rectangle, which is an instance of the second quantized electronic structure Hamiltonian discretized using Gaussian basis functions. We study this problem in a minimal basis set, which exhibits false minima that can trap the standard variational algorithm despite the problem's small size. We show that augmenting the training with parallel tempering becomes useful to finding good approximations to the ground states of these problem instances.

28th March 2023 | 08:00 UTC | leva Cepaite | University of Strathclyde

Counterdiabatic Optimised Local Driving

Counterdiabatic driving protocols provide a way to speed up adiabatic dynamics through the suppression of excitations into unwanted states. However, exact counterdiabatic drives require knowledge of the spectral properties of the instantaneous Hamiltonians, which limits their application. It has recently been shown that this requirement can be removed by using a variational approach which allows one to determine an approximate, localised counterdiabatic drive without requiring access to the wavefunction of the system. We show that when coupled with additional parameterised driving terms, the path of the system can be optimised in a way that enhances the loss-suppressing effects of a chosen local order of the counterdiabatic protocol. Furthermore, this path appears to minimise the total power and maximal driving strength of higher local order terms in the approximate counterdiabatic drive. We show that this can be used in order to perform optimisation of the system path in a way that suppresses non-diabatic losses without requiring access to the system dynamics or experimental results.

4th April 2023 | 16:00 UTC | Matteo Michele Wauters | University of Copenhagen

Optimizing annealing paths with Monte Carlo tree search

Schedule optimization in quantum annealing processes is a crucial ingredient in improving the performance of adiabatic algorithms in ground state preparation. Among other gradient-free approaches, Monte Carlo tree search (MCTS) algorithms have been successfully applied in the strategy design for complex games such as chess or go [1]. Expanding on the recent work of Chen et al.[2], we investigate the advantages and limitations of MCTS approaches as annealing schedules optimizers and compare them with standard gradient-based methods. We test these ideas on hard 3-SAT instances and Max-Cut problems on regular graphs, showing that MCTS has a similar performance of gradient descent methods regarding the algorithm accuracy while showing indications of a better scaling of the computational cost with the number of variational schedule parameters. We also argue that it might lead to improved robustness against hardware and measurement noise.

12th April 2023 | 00:01 UTC | Alexander Whiticar | D-Wave Systems

Probing flux and charge noise with macroscopic resonant tunneling

This talk will report on measurements of flux and charge noise in an rf-SQUID flux qubit using macroscopic resonant tunneling (MRT). We measure rates of incoherent tunneling from the lowest energy state in the initial well to the ground and first excited states in the target well. The result of the measurement consists of two peaks. The first peak corresponds to tunneling to the ground state of the target well, and is dominated by flux noise. The second peak is due to tunneling to the excited state and is wider due to an intrawell relaxation process dominated by charge noise. We develop a theoretical model that allows us to extract information about flux and charge noise within one experimental setup. The model agrees very well with experimental data over a wide dynamic range and provides parameters that characterize charge and flux noise.

18th April 2023 | 08:00 UTC | Hiroshi Hayasaka | AIST

Adiabatic condition for quantum annealing revisited

Quantum annealing (QA) is metaheuristics of combinatorial optimization problem using quantum fluctuation. If the adiabatic condition is satisfied in the QA, the ground state of the problem Hamiltonian can be obtained. The adiabatic condition consists of a transition matrix of time derivative of Hamiltonian and an energy gap between the ground and excited states. It is believed that the scaling of the energy gap provides computational complexity of the combinatorial optimization problems. In this presentation, we propose a general framework that gives counterintuitive models to this common wisdom: QA with a constant annealing time fails despite a constant energy gap, i.e., O(L^0) during QA, where L is the problem size. The key idea of our analysis is to add a penalty term in the Hamiltonian, which does not change the eigenstate of Hamiltonian but change the eigenvalue. We investigate the adiabatic Grover search with penalty term as the concrete example, we analytically show that the transition matrix becomes exponentially large and first-order phase transition occurs despite the energy gap that scales as O(L^0). Moreover, we show that, in this example, the success probability of QA becomes exponentially small as we increase the problem size L. This paper was based on results obtained from a project, JPNP16007, commissioned by the New Energy and Industrial Technology Development Organization (NEDO), Japan.

25 April 2023 | 16:00 UTC | Jeremy Cote | Université de Sherbrooke

Quantum Annealing Without Wasting Time

In quantum annealing, we use a schedule to interpolate between a driver Hamiltonian with an easy-to-prepare ground state and a problem Hamiltonian whose ground state encodes solutions to a classical optimization problem. The standard implementation relies on the evolution being adiabatic: keeping the system in the instantaneous ground state with high probability and requiring a time scale inversely related to the minimum energy gap between the instantaneous ground and excited states. For adiabatic evolution, this time can scale exponentially with the system size, even for computationally simple problems. But we don't actually care about the evolution being adiabatic; what we care about is measuring solutions to the problem at the end of the protocol. By optimizing the annealing schedules, can we avoid the exponential slow down that often occurs with standard quantum annealing? In this presentation, I'll share our investigation of this question for a class of problems called the frustrated ring model.

3 May 2023 | 00:01 UTC | Elijah Pelofske | Los Alamos National Laboratory

Quantum Annealing vs. QAOA: 127 Qubit Higher-Order Ising Problems on NISQ Computers

Quantum annealing (QA) and Quantum Alternating Operator Ansatz (QAOA) are both heuristic quantum algorithms intended for sampling optimal solutions of combinatorial optimization problems. In this article we implement a rigorous direct comparison between QA on D-Wave hardware and QAOA on IBMQ hardware. The studied problems are instances of a class of Ising problems, with variable assignments of +1 or -1, that contain cubic ZZZ interactions (higher order terms) and match both the native connectivity of the Pegasus topology D-Wave chips and the heavy hexagonal lattice of the IBMQ chips. The novel QAOA implementation on the heavy hexagonal lattice has a CNOT depth of 6 per round and allows for usage of an entire heavy hexagonal lattice. Experimentally, QAOA is executed on an ensemble of randomly generated Ising instances with a grid search over 1 and 2 round angles using all 127 programmable superconducting transmon qubits of ibm_washington. The error suppression technique digital dynamical decoupling (DDD) is also tested on all QAOA circuits. QA is executed on the same Ising instances with the programmable superconducting flux qubit devices D-Wave Advantage_system4.1 and Advantage_system6.1 using modified annealing schedules with pauses. We find that QA outperforms QAOA on all problem instances. We also find that dynamical decoupling enables 2-round QAOA to outperform 1-round QAOA.

9 May 2023 | 08:00 UTC | Vikrant Kumar | TCG CREST

Electronic Structure Calculations using Quantum Annealing

Quantum annealing of Ising Hamiltonians is a powerful heuristic for solving quadratic binary optimization problems. The recently proposed Quantum Annealer Eigensolver algorithm has leveraged this idea to estimate the eigenvalues of electronic structure Hamiltonians. However, the limited spin interactions in current D-Wave devices require that more complex problems with denser interactions need to be transformed into the sparse Ising spin form, which introduces computation errors. To overcome this, we present a hybrid quantum-classical protocol that features a new scheme for mapping continuous variable functions to binary and a perturbation theory-inspired divide-and-conquer approach. Our approach has yielded promising results for computing the relativistic effects in heavy ions using quantum annealers. We will also compare our method's efficiency against previously implemented heuristics for combinatorial optimization problems.

16 May 2023 | 16:00 UTC | Gianluca Passarelli | SPIN (CNR)

Counterdiabatic reverse annealing

In this talk, I will discuss counterdiabatic driving and reverse annealing, two quantum techniques that play important roles in suppressing nonadiabatic effects and finding the ground state of a system, respectively. I will then introduce counterdiabatic reverse annealing, a quantum annealing protocol that combines the advantages of these two techniques to extend the range of reverse annealing to short-time domains. This technique utilizes approximate counterdiabatic driving through low-order nested commutators and has the potential for near-term experimental implementation in quantum devices. I will compare the performance of counterdiabatic reverse annealing to unassisted reverse annealing using metrics such as ground-state fidelity and time to solution.

24 May 2023 | 00:01 UTC | Shyam Dhamapurkar | Southern University of Science & Technology

Quantum walks as thermalizations, with application to fullerene graphs

We consider whether quantum walks can constitute models of thermalization, analogously to how classical random walks can be models for classical thermalization. In a quantum walk over a graph, a walker moves in a superposition of node positions via a unitary time evolution. We show a quantum walk can be interpreted as an equilibration of a kind investigated in the literature on thermalization in unitarily evolving quantum systems. This connection implies that recent results concerning the equilibration of observables can be applied to analyse the node position statistics of quantum walks. We illustrate this in the case of a family of graphs known as fullerenes. We find that a bound from Short et al., implying that certain expectation values will at most times be close to their time-averaged value, applies tightly to the node position probabilities. Nevertheless, the node position statistics do not thermalize in the standard sense. In particular, quantum walks over fullerene graphs constitute a counter-example to the hypothesis that subsystems equilibrate to the Gibbs state. We also show how quantum walks can be used to probe the universality of the ETH relation. We find that in C60 the relation does not hold for node position projectors, but it does hold for the average position. The results create a concrete bridge between the study of self-thermalization of quantum systems and that of quantum computation via quantum walks.

30 May 2023 | 08:00 UTC | Merlin Nau | University of Erlangen-Nuremberg

Hybrid quantum annealing for tomographic image reconstruction - opportunities and limitations

We discuss the potential applications of quantum annealing in tomographic image reconstruction and medical imaging. Our focus is on the challenges of reconstructing tomographic images with limited measurements and a low signal-to-noise ratio, a critical issue in clinical imaging for improving patient comfort and reducing radiation exposure. We propose using a quantum annealer and associated hybrid methods to solve this reconstruction problem, as quantum computing continues to advance. We present the results of our QA-based reconstruction technique, which we tested for image size, noise content, and underdetermination of the measured projection data. Our reconstructed binary and integer-valued images compete with traditional reconstruction algorithms and are superior in terms of noise robustness and reconstruction from few projections, for small scale images.

06 June 2023 | 16:00 UTC | Aaron Villanueva | Radboud University Nijmegen

Grover-like speedup for combinatorial optimization and the fine-tuning problem in quantum annealing

We study a quantum annealing model for combinatorial optimization problems with Hamiltonian $H = H_0 + z H_f$ where $H_f$ is diagonal, $H_0=-\ket{\phi}\bra{\phi}$ is the equal superposition state projector and $z$ the annealing parameter. We analytically compute the minimal spectral gap, which is $\Omega(1/\sqrt{N})$, and its location. We show that quantum speed-up requires an annealing schedule which demands a precise knowledge of the minimal gap location, which can be computed only if the density of states of the optimization problem is known. Assuming that the density of states is known, we prove the existence and optimality of quadratic speed-up. However, in general the density of states is intractable to compute, making quadratic speed-up unfeasible for any practical optimization problems. We conjecture that it is likely that this negative result also applies for any other instance independent transverse Hamiltonians such as $H_0 = -\sum_{i=1}^n \sigma_i^x$.

14 June 2023 | 00:01 UTC | David Haycraft | Quantum Computing Inc

Suppressing unwanted fluctuations in QAOA and approximate quantum annealing

The quantum approximate optimisation algorithm (QAOA) was partially inspired by digitising quantum annealing. Based on this inspiration, we develop techniques to use the additional flexibility of a universal gate-model quantum computer to mitigate fluctuation effects which are known to distort the search space within quantum annealing and lead to false minima. We find that even just the added ability to take Pauli X measurements allows us to modify the mixer angles to counteract these effects by scaling mixer terms in a way proportional to the diagonal elements of the Fubini-Study metric. We find that mitigating these effects can lead to higher success probabilities in cases where the energy landscape is distorted and that we can use the same Pauli X measurements to target which variables are likely to be susceptible to strong fluctuations. The effects of the methods we introduce are relevant even at relatively low depth of p = 10 −20, suggesting that the techniques we are developing are likely to be relevant in the near term. Furthermore, since these methods rely on controlling a degree of freedom which is not typically modified in QAOA, our methods will be compatible with a wide range of other QAOA innovations. We further verify that these fluctuation effects can be observed on an IonQ Harmony QPU.

27 June 2023 | 16:00 UTC | Enrico Zardini | University of Trento

Quantum Annealing Learning Search Implementations

This talk presents the details and testing of two implementations (in C++ and Python) of the hybrid quantum-classical algorithm Quantum Annealing Learning Search (QALS) on a D-Wave quantum annealer. QALS was proposed in 2019 as a novel technique to solve general QUBO problems that cannot be directly represented into the hardware architecture of a D-Wave machine. Repeated calls to the quantum machine within a classical iterative structure and a related convergence proof originate a learning mechanism to find an encoding of a given problem into the quantum architecture. In this talk, we consider the Number Partitioning Problem (NPP) and the Travelling Salesman Problem (TSP) for the testing of QALS. The results turn out to be quite unexpected, with QALS not being able to perform as well as the other considered methods, especially in NPP, where classical methods outperform quantum annealing in general. Nevertheless, looking at the TSP tests, QALS has fulfilled its primary goal, i.e., processing QUBO problems not directly mappable to the QPU topology.

11 July 2023 | 08:00 UTC | Kaho Takahashi | Hitachi

Overview and Application Example of CMOS Annealing

Hitachi has developed a unique computing technology called CMOS Annealing. It is specialized in solving a combinatorial optimization problem formulated as a quadratic unconstrained binary optimization (QUBO) problem. In this talk, we will present an overview of CMOS Annealing and our experimental results on relief supply planning. In the case of a large-scale disaster, it is necessary to optimize a large-scale distribution network at high speed, and also to distribute supplies evenly to all areas, as opposed to general delivery plans. We formulated a QUBO using quadratic terms to express the equality of supplies within the affected area. We applied CMOS annealing to a relief supply delivery plan for the Tokyo metropolitan area, and it is confirmed that the total distance traveled can be reduced and delivery routes with higher equality can be calculated compared to conventional rule-based calculation results.

18 July 2023 | 16:00 UTC | Cathy McGeoch | D-Wave Systems

Quantum Utility and Quantum Performance Mechanisms

Quantum utility is a proposed new benchmark for evaluating quantum performance as experienced by the user. It considers the quality of the quantum computation together with overhead costs, asking: can the quantum processor outperform classical solvers at some tasks of interest to practitioners, when considering computational overheads? A milestone is a limited form of quantum utility that focuses on restricted subsets of overheads. We describe performance of a D-Wave Advantage QPU with respect to three milestones: M0 measures pure quantum computation time (no overheads); M1 measures access time (including programming and readout); and M2 measures an indirect cost associated with minor embedding. The QPU outperforms 7 classical solvers on 13 inputs in 99% of tests under M0 and M1, and in about 19% of tests under M2. Comparison of test results under M2 reveals some striking differences between quantum and classical performance mechanisms, highlighting distinctions between ``hard’’ or ``easy’’ inputs for each paradigm. These distinctions bode well for future annealing QPUs to support demonstrations of quantum utility on broader classes of inputs and on more challenging milestones

26 July 2023 | 00:01 UTC | Nikolai Sinitsyn | Los Alamos National Laboratory

Adiabatic oracle for Grover Algorithm

There is a hidden belief among many physicists that even perfect quantum annealing computers cannot solve complex constraint-satisfaction problems faster than classically. The reason is that the mathematically proved algorithms use certain resources that are assumed to be given at almost no cost but are hard to provide in practice. An alternative is to work with “heuristic” approaches, such as DWave-like quantum annealing but there is no evidence that it works without similar “oracle”-like assumptions at some stage [1]. In this talk I will argue that practically useful and un-classically fast quantum computing is possible without assumptions that something is given for free. Namely, I will describe a “hybrid” approach that realizes Grover’s oracle for numerous computational problems by means of quantum annealing in polynomial time [2-3].

6 September 2023 | 00:01 UTC | Michael Huang | University of Rochester

Dynamical-System Computing Platform: A case study on BRIM

In recent years, a growing number of novel systems based on diverse physics have demonstrated strong computing capabilities for special types of problems such as QUBO (or equivalently Ising model optimization) problems. These machines are often loosely termed Ising machines. While CMOS technology has been predominantly used to build conventional computing systems, there is no fundamental reason why it cannot be used to build similar dynamical systems that can solve QUBO or similar problems with significant efficiency benefits. In this talk, I will discuss one such incarnation called BRIM recently developed at the University of Rochester. I will show the working principle as well as estimated performance parameters. I will also show some of the benefits of building such CMOS-compatible dynamical systems with an example multiple-processor architecture that scales the capacity of the Ising machine and customization to more efficiently solve SAT problems. Overall, I hope to make the point that such architectures have a place in the future computational infrastructure.

12 September 2023 | 08:00 UTC | Jessica Park | University of York

Spatial Correlations in the Qubit Properties of Measured and Simulated Spin Networks

We show strong positive spatial correlations in the qubits of a D-Wave 2000Q quantum annealing chip that are connected to qubits outside their own unit cell. By simulating the dynamics of spin networks, we then show that correlation between nodes is affected by a number of factors. The different connectivity of qubits within the network means that information transfer is not straightforward even when all the qubit-qubit couplings have equal weighting. The similarity between connected nodes is further changed when the couplings' strength is scaled according to the physical length of the connections (here to simulate dipole-dipole interactions). This highlights the importance of understanding the architectural features and potentially unprogrammed interactions/connections that can divert the performance of a quantum system away from the idealised model of identical qubits and couplings across the chip.

19 September 2023 | 16:00 UTC | Rudi Finzgar | Technical University of Munich

Designing Quantum Annealing Schedules using Bayesian Optimization

We propose and analyze the use of Bayesian optimization techniques to design quantum annealing schedules with minimal user and resource requirements. We showcase our scheme with results for two paradigmatic spin models. We find that Bayesian optimization is able to identify schedules resulting in fidelities several orders of magnitude better than standard protocols for both quantum and reverse annealing, as applied to the p-spin model. We also show that our scheme can help improve the design of hybrid quantum algorithms for hard combinatorial optimization problems, such as the maximum independent set problem, and illustrate these results via experiments on a neutral atom quantum processor available on Amazon Braket.

27 September 2023 | 00:01 UTC | Hiroshi Hayasaka | AIST

A general method to construct mean field counter diabatic driving for quantum annealing

Quantum annealing (QA) has been attracted much attention for exploring the ground states of quantum many-body systems. While QA typically requires slow dynamics to satisfy adiabatic conditions, counter-diabatic (CD) driving can suppress non-adiabatic transitions, enabling fast QA [1, 2]. However, constructing the CD term necessitates exact eigenstate computation via classical computer and poses experimental challenges due to its non-local nature.

To address this, we propose a practical method for approximating the CD term using a mean-field (MF) approach. In our approach, solving self-consistent equations at each time step is unnecessary; only the initial configuration is required. Numerical results show that dynamics with the MF-approximated CD term replicate self-consistent solutions. Applying this method to quantum spin glasses improves ground state fidelity compared to traditional QA. Furthermore, we experimentally validate our approach using a D-wave quantum annealer, Advantage, obtaining results that support our numerical simulations [3]. This paper was based on results obtained from a project, JPNP16007, commissioned by the New Energy and Industrial Technology Development Organization (NEDO), Japan.

[1] M. Demirplak, et al., J. Phys. Chem. A 107, 9937 (2003)

[2] M. V. Berry, J. Phys. A: Mathematical and Theoretical 42, 365303 (2009)

[3] H. Hayasaka, et al., arXiv: 2305.08352

10 October 2023 | 16:00 UTC | Federico Balducci | University of Luxembourg

Large Deviations Beyond the Kibble-Zurek Mechanism

Crossing a quantum phase transition in finite time leads to the formation of excitations, such as topological defects, since the dynamics necessarily fails to be adiabatic near the critical point. The average number of excitations is well described by the celebrated Kibble-Zurek (KZ) mechanism, predicting a universal scaling law with the quench time. Recently, the scope of the KZ paradigm has been expanded, enabling the prediction of quantities beyond averages, such as the full counting statistics of defects [1]. In this talk, I will present some results [2] that clarify the role of universality in beyond-KZ physics, by borrowing tools from Large Deviations Theory. Using the transverse-field Ising model as test bed, I will show how the rate function obeys a universal scaling relation with the quench time. Then, I will expand the result to classical phase transitions, using few additional assumptions on the way defects form. I will finally argue how these theoretical predictions are already testable in current quantum simulators and annealers [3].

REFS:

[1] Gómez-Ruiz et al, Phys. Rev. Lett. 124, 240602 (2020)

[2] Balducci et al, arXiv:2307.02524

[3] Bando et al, Phys. Rev. Research 2, 033369 (2020)

24 October 2023 | 08:00 UTC | Sebastian Zielinski | Ludwig Maximilian University of Munich

The Model Matters – Influence of QUBO models on the solution quality of quantum annealing

To solve 3SAT instances on quantum annealers they need to be transformed to instances of Quadratic Unconstrained Binary Optimization (QUBO). When there are multiple transformations available, the question arises whether different transformations lead to differences in the obtained solution quality. In this presentation we will see that the choice of a 3SAT-to-QUBO transformation can significantly impact the number of correct solutions that a quantum annealer returns. Furthermore, we will see that the size of a QUBO instance (i.e., the dimension of the QUBO matrix) is not a sufficient predictor for the solution quality, as larger QUBO instances may produce better results than smaller QUBO instances for the same problem. Motivated by these results, we will learn about the concept of “Pattern QUBOs” and show how they can be used to create (millions of) new 3SAT-to-QUBO transformations automatically via a novel algorithmic method. Furthermore the formerly mentioned procedure also enables easier access to quantum technologies, as it is now no longer needed to deeply understand 3SAT-to-QUBO transformations, because these transformations can no be created fully automatically.

31 October 2023 | 17:00 UTC | Michael Huang | University of Rochester

Dynamical-System Computing Platform: A case study on BRIM

14 November 2023 | 09:00 UTC | Alejandro Montanez-Barrera | Forschungszentrum Jülich University

Improving Performance in Combinatorial Optimization Problems with Inequality Constraints: An Evaluation of the Unbalanced Penalization Method on D-Wave Advantage

Combinatorial optimization problems are one of the target applications of current quantum technology, mainly because of their industrial relevance, the difficulty of solving large instances of them classically, and their equivalence to Ising Hamiltonians using the quadratic unconstrained binary optimization (QUBO) formulation. Many of these applications have inequality constraints, usually encoded as penalization terms in the QUBO formulation using additional variables known as slack variables. The slack variables have two disadvantages: (i) these variables extend the search space of optimal and suboptimal solutions, and (ii) the variables add extra qubits and connections to the quantum algorithm. Recently, a new method known as unbalanced penalization has been presented to avoid using slack variables. This method offers a trade-off between additional slack variables to ensure that the optimal solution is given by the ground state of the Ising Hamiltonian, and using an unbalanced heuristic function to penalize the region where the inequality constraint is violated with the only certainty that the optimal solution will be in the vicinity of the ground state. This work tests the unbalanced penalization method using real quantum hardware on D-Wave Advantage for the traveling salesman problem (TSP). The results show that the unbalanced penalization method outperforms the solutions found using slack variables and sets a new record for the largest TSP solved with quantum technology.

21 November 2023 | 17:00 UTC | Lorenzo Rocutto | Italian Institute of Technology

Performance-Improving Techniques for Adiabatic Quantum Computers

We challenged a D-Wave device on two industrially-relevant computational problems to evaluate the dependence of its performances on the internal parameters used during annealing. In one of the two tested problems, a simple parameter tuning reduced the computational time required to find the global optimum by 78 times. Surprisingly, the application of parallel computing techniques can significantly further reduce the computational cost. We show novel, unpublished results regarding the implementation of a fully-connected BM on a D-Wave device where we observed a limited quantum advantage. The quantum approach was compared to a classical, Metropolis-based approach implemented on GPUs. Such results, corroborated by theoretical arguments, suggest that the first useful application for Adiabatic Quantum Computers could be in sampling tasks, despite this technology has often been advertised as an optimization problems solver.

29 November 2023 | 01:00 UTC | Samyak Jhaveri | University of California, Irvine

Cloning and Beyond: A Quantum Solution to Duplicate Code

Quantum computers are becoming a reality. The advantage of quantum computing is that it has the potential to solve computationally complex problems in a fixed amount of time, independent of the size of the problem. However, the kinds of problems for which these computers are a good fit, and the ways to express those problems, are substantially different from the kinds of problems and expressions used in classical computing. Quantum annealers, in particular, are currently the most promising and available quantum computing devices in the short term. However, they are also the most foreign compared to classical programs, as they require a different kind of computational thinking. In order to ease the transition into this new world of quantum computing, we present a novel quantum approach to a well-known software problem: code clone detection. We express code clone detection as a subgraph isomorphism problem that is mapped into a quadratic optimization problem, and solve it using a DWave quantum annealing computer. We developed a quantum annealing algorithm that compares Abstract Syntax Trees (AST) and reports an energy value that indicates how similar they are. The motivation behind this research goes well beyond code duplicate detection: our approach paves the way into how to express software engineering problems as optimization problems that can be solved by quantum annealers.

5 December 2023 | 09:00 UTC | Takashi Imoto | AIST

Measurement of the incoherent decay rate of quantum states in quantum annealing with a D-Wave machine

Quantum annealing has been demonstrated with superconducting qubits. Such a quantum annealer has been used to solve combinational optimization problems and is also useful as a quantum simulator to investigate the properties of the quantum many-body systems. However, the coherence properties of actual devices provided by D-Wave Quantum Inc. are not sufficiently explored. Here, we propose and demonstrate a method to measure the coherence time of the excited state in quantum annealing with the D-Wave device. More specifically, we investigate the energy relaxation time of the first excited states of a fully connected Ising model with a transverse field. We find that the energy relaxation time of the excited states of the model is orders of magnitude longer than that of the excited state of a single qubit, and we qualitatively explain this phenomenon by using a theoretical model. The reported technique provides new possibilities to explore the decoherence properties of quantum many-body systems with the D-Wave machine.

12 December 2023 | 17:00 UTC | Marcin Dukalski | Aramco Overseas Company

Imaging shallow subsurface with quantum annealers

Imaging science attempts to determine the properties of the medium from indirect measurements, e.g. wave scattering. Subsurface, and in particular seismic, imaging is probably one of the most computationally demanding examples thereof. It involves finding solutions to a number of very large optimization problems characterized by severely multi-modal objective functions. In this talk we wish to showcase our progress on a geophysical problem which fits the quantum annealing paradigm. On land, where geological anomalies such as underground rivers, caves, or boulders, create abrupt local misalignment in measured wavefronts, which once corrected result in sharper image and greater confidence in the imaged geological formations. Classical methods either resort to a mean field type approach or are susceptible to returning potentially far away in parameter space local optima, which might give a false sense of certainty about the subsurface. Solving this problem has far reaching implications to the economics of civil engineering, hydrology, CO2 or H2 gas storage and hydrocarbon exploration. We reformulate the problem as a Potts model and show how one address it with a classical-quantum hybrid solver. We show how we produce shallow subsurface maps based on processing of gigabytes of seismic data.