Session H06: Quantum Computation: Gates, Algorithms, & Architectures

Live

The rise of commercial quantum computing devices has allowed for the testing of small-scale quantum algorithms on physical systems. These devices are generally limited in the flexibility of type, ordering and timing of operations, making them restrictive for researchers developing novel quantum protocols. In this talk we present progress towards a trapped ion system designed to have greater flexibility in order to provide a useful tool for the academic research community. Researchers will have remote access to native light-matter interactions that drive trapped ion quantum gates and therefore have the greatest possible suite of operations for realizing novel quantum computing protocols. This functionality is provided by a custom control system based on commercial FPGA boards to realize autonomous running of the trapped ion system with sub-nanosecond timing precision. Initially, the system will allow for fully connected interactions between up-to 16 individually addressed ions, with a modular design allowing for future upgrades to the system. The individual addressing scheme is based on a guided-light platform, enabled by the 532 nm wavelength of Ba$^+$ Raman transitions. We will present the progress made in the building of the system, and the challenges that lie ahead.   

Live

Trapped-ion entangling gates typically rely on laser-induced coupling of the ions’ internal qubit states to their motion. Laser-free entangling gates, which eliminate photon scattering errors and may offer benefits for scalability, have been implemented using either static magnetic field gradients or magnetic field gradients oscillating at GHz frequencies. We demonstrate a recently proposed trapped-ion entangling gate based on a radiofrequency oscillating magnetic field gradient and two microwave magnetic fields symmetrically detuned about the qubit frequency [1]. Our implementation offers reduced sensitivity to qubit and motional frequency errors, enabling the generation of a symmetric Bell state of two $^{25}$Mg$^{+}$ hyperfine qubits with fidelity $0.998(2)$ in $700$ $\mu$s. This gate scheme also allows us to incorporate laser-free single-ion addressing to prepare an antisymmetric Bell state. [1] R.T. Sutherland et al., New J Phys 21, 033033 (2019)   

Live

Quantum-classical hybrid schemes, such as the Quantum Approximate Optimization Algorithm (QAOA), are promising approaches to solving combinatorial optimization problems on near-term quantum hardware. Counting edge cover, which is a set of edges that leaves no isolated vertices on a graph, is one of these problems and has important applications in network reliability. We implement a modified QAOA scheme with an adapted mixing Hamiltonian on an ion trap quantum computer to count edge covers of a weighted 3-node star graph. This modified algorithm prepares the quantum system in a superposition of ground states with pre-determined weights for efficient counting. We demonstrate how the approximate solution becomes more exact with increasing number of QAOA-layers on real hardware, despite the additional gate errors.   

Live

The unitary coupled-cluster ansatz is simple to implement in a factorized form on quantum computers. Instead of focusing on variational quantum eigensolvers in this work, we focus on using the unitary coupled-cluster as a tool for ground-state preparation to enable further quantum computation. As an example, we prepare the ground state of a Hydrogen molecule in a minimal basis and illustrate how to use it to determine the "vertical" ionization potential. We deploy this approach on current IBM hardware and discuss how well it works on NISQ machines. If available at the time of the talk, we will also illustrate results for larger basis sets too.   

Live

We report the discovery of a conceptually simple technique to generate and tailor universal quantum computing resources known as cluster entangled states, here of light. Our method uses a single source of pairwise entangled optical fields, i.e., an optical parametric oscillator, and an electro-optic modulator, a common photonic device. We show that the combined action of the quantum source and of the phase modulator involves several tunable parameters that confer a high degree of quantum control over the generated quantum state, and can increase its topological dimension. This extremely simple architecture is highly compatible with on-chip integrated optics.   

Live

Quantum error correction is crucial for constructing a fault-tolerant quantum computer. By employing redundancy, error-correcting codes protect logical qubits against errors at the physical-qubit-level during state preparation, operations and measurement. Here we demonstrate an encoding of a logical qubit with the Shor code, which detects and corrects single-qubit bit-flip and phase-flip errors, on a trapped ion system. Using nine physical qubits, we prepare a logical state $\ket{0}$ with $98.75\%$ fidelity and a logical state $\ket{1}$ with $98.51\%$ fidelity after correction with majority voting. We further investigate the robustness of the logical qubit and shows data extrapolating its performance to deeper encodings.   

On Demand

We discuss new ideas for quantum logic using both dipole-dipole and dipole-phonon interactions between sympathetically cooled molecular ions and summarize our experimental efforts in realizing these quantum logic schemes. Rather than coupling qubits by their shared motion, we explore the prospect of using the dipole-dipole interaction between molecular ions. A static polarizing electric field can facilitate this dipole-dipole interaction in neutrals, but in an ion trap the time-averaged electric field experienced by an ion is zero. Instead, by creating a superposition of opposite-parity eigenstates, the time-dependent polarization mediates a dipole-dipole interaction. The use of such oscillating dipole moments dynamically decouples the dipoles from laboratory electric fields, including those of the ion trap. As such, this technique is relatively insensitive to anomalous heating in ion traps. Additionally, we consider the coupling of a dipole moment of a polar molecular ion with the phonon modes of a Coulomb crystal. When the transition frequency between two dipole states is similar to the normal mode frequencies in an ion trap, the interaction between dipole and phonons become important. This interaction can be utilized in a number of promising applications.   

A Midscale Quantum Computer Based on Trapped Ions

In collaboration between universities and industrial partners, we have constructed an trapped ion-based quantum computer with the goal of realizing an error-corrected quantum bit and alternatively algorithms that do not need error correction. We report on the performance of our system, including fidelities of single-qubit and two-qubit gates and its operation with long and mixed-species ion chains. We present the building blocks of a logical qubit in our system, including qubit encoding, operation of logical gates, and stabilizer measurements, and discuss the progress towards implementing the full error correction algorithm as well as other applications.   

Towards robust two-qubit gates on a trapped-ion quantum computer

The ability to implement robust entangling gates on a quantum computer is essential to scalable quantum computing. In this talk, I will present a constructive method to shape pulses that implement two-qubit XX gates on a trapped-ion quantum computer. By modulating amplitude and frequency of the pulses that illuminate the ions simultaneously, this method can stabilize the XX gates against external parameter fluctuations. The method is linear, requiring only modest amount of classical computational resources.   

Training Optimization for Gate-Model Quantum Neural Networks

Gate-based quantum computations represent an essential to realize near-term quantum computer architectures. A gate-model quantum neural network (QNN) is a QNN implemented on a gate-model quantum computer, realized via a set of unitaries with associated gate parameters. Here, we define a training optimization procedure for gate-model QNNs. By deriving the environmental attributes of the gate-model quantum network, we prove the constraint-based learning models. We show that the optimal learning procedures are different if side information is available in different directions, and if side information is accessible about the previous running sequences of the gate-model QNN. The results are particularly convenient for gate-model quantum computer implementations.