BeQuEST: Benchmarking Quantum Enhancement for Science & Technology
Sponsors – DARPA
The Quantum Benchmarking program will estimate the long-term utility of quantum computers by creating new benchmarks that quantitatively measure progress towards specific, transformational computational challenges. In parallel, the program will estimate the hardware-specific resources required to achieve different levels of benchmark performance.
Towards Enhanced Quantum Annealing in Learning and Simulation (PI: Daniel Lidar)
Sponsors – DARPA
Quantum annealing was the first quantum computing methodology to be commercialized, and it continues to lead in practical applications of quantum hardware-based algorithms. However, to achieve real-world usefulness, quantum annealing will need to overcome three obstacles: (1) higher coherence, (2) compatibility with application problems, and (3) an application domain with a demonstrable quantum advantage. This project, “TEQUILAS: Towards Enhanced Quantum annealing In Learning And Simulation,” will develop methods to address the embedding of fully connected logical problems in hardware and applications in the domain of machine learning using quantum Boltzmann machines and quantum simulation.
Resource Efficient Quantum Simulations on NISQ Devices: Advancing the State of the Art
Sponsors – DOE
Simulating quantum many-body systems is a central challenge in Physics, Chemistry and Material Sciences, and other areas, aiming to advance our understanding of phenomena associated with controlling the quantum dynamics of non-equilibrium chemical and materials systems, unraveling the physics and chemistry of strongly correlated electron systems, and more. Although no classical algorithms efficiently simulate quantum many-body systems, quantum algorithms offer a way around the classical bottlenecks by “digitizing” the time evolution of the system in question on a circuit. We will theoretically develop and experimentally benchmark a novel algorithmic approach for resource-efficient Hamiltonian simulations that will supersede the current state-of-the-art. Our technique utilizes aspects of quantum Hamiltonian simulation not taken into account in existing Hamiltonian simulation algorithms, specifically, prior knowledge of the structure of the Hamiltonian and its relation to the initial state of the system, in order to further reduce the resource cost of the simulation. We will show that this additional information allows for a series expansion of the quantum time-evolution operator that takes advantage of this previously disregarded information. The advantages of the new approach translate to a more efficient, or resource-lean, Hamiltonian simulation than the existing state-of-the-art. In addition, this effort will have an important experimental component whereby the proposed technique will be benchmarked on currently available quantum devices, and the capabilities–as well as the limitations–of noisy intermediate-scale quantum (NISQ) hardware will be determined.
Software Stack and Algorithms for Automating Quantum Classical Computing
Sponsors – DOE
We will develop an open-source algorithm and software stack that will automate the process of designing, executing, and analyzing the results of quantum algorithms, thus enabling new discovery across scientific domains. Prompted by limitations on algorithms and models of computation imposed by available and near-term quantum hardware – noisy quantum gates, limited qubit connectivity, and short circuit depth to name a few – we focus the stack on implementing a hybrid quantum-classical model of computation for various types of quantum hardware with particular emphasis on scientific applications in quantum field theory, nuclear physics, condensed matter, and quantum machine learning.
Q4Q: Quantum Computation for Quantum Prediction of Materials and Molecular Properties (PI: Rosa Di Felice)
Sponsors – DOE
Computational studies have boosted knowledge and progress in chemistry and materials science, by enabling prediction of novel phenomena in molecules and materials, which is at the core of DOE’s mission. While computers have undergone enormous progress in the last 3 decades, the underlying silicon technology is coming to a critical stage and the scientific community is looking for alternatives. Quantum computation (QC) is a potential option. Quantum information theory has made great progress in producing a mature formulation of concepts, theorems and algorithms. Yet, the field of quantum computation has been frustrated by the difficulty of fabricating useful hardware. Two other significant issues for the development of quantum computing have been: the hindrance to identify breakthrough applications and the difficulty of mastering both quantum computer science and chemistry/materials science. However, these hurdles are beginning to be overcome. The two research areas identified for their potential synergy with quantum hardware and software targeted in this project are “controlling the quantum dynamics of nonequilibrium chemical and materials systems” and “unraveling the physics and chemistry of strongly correlated systems”. The quantum devices that are currently accessible to users implement two different schemes for quantum computation: (a) gate‐model quantum computation (GMQC) in IBM and Rigetti cloud hardware and (b) adiabatic quantum optimization (AQO) in D‐Wave commercial hardware. We are using all these platforms to solve simple problems in chemistry and materials science.
QAFS: Quantum Annealing Feasibility Study (PI: Daniel Lidar)
Sponsors – DARPA
QAFS seeks to harness quantum effects required to enhance quantum annealing solutions to hard combinatorial optimization problems. The physics underlying quantum enhancement will be corroborated by design and demonstration of research-scale annealing test beds comprised of novel superconducting qubits, architectures, and operating procedures. All work will serve to demonstrate a plausible path to enhancement and a basis for design of application-scale quantum annealers.