EFFICIENT NUMERICAL METHODS FOR STUDYING THE COMPLEXITY VS. ROBUSTNESS TRADEOFF IN NOISY INTERMEDIATE-SCALE QUANTUM (NISQ) INFORMATION PROCESSORS
Abstract
Quantum information science represents a radical departure from the paradigm of classical information. By employing a new class of devices that take advantage of the laws of superposition and entanglement in quantum mechanics, quantum information processors could outperform their classical counterparts and enable the exploration of new phenomena that have been out of reach because of the limitations of even the fastest supercomputers that fundamentally process classical information. Of particular relevance is the task of quantum simulation, whereby static or dynamical properties of quantum systems could be explored by simulating them on a quantum information processor. Such a quantum simulator would in principle allow us to address important questions about the nature of highly correlated systems, such as fermionic systems and frustrated spin systems, at the heart of many important problems in chemistry and physics, that have been so far inaccessible by classical simulation. This promise has fueled e orts to control, manipulate, and measure quantum systems, and the progress made on these fronts has heralded the arrival of multiple platforms for quantum information processors. While representing a significant achievement, these devices remain limited in their capabilities; they do not implement quantum error correction, so they remain susceptible to the loss of quantum information and to finite implementation precision. Such sources of error inevitably limits the complexity of quantum states that can be prepared by the quantum simulator, which in turn limits the possible quantum advantage that can be achieved by such a device. Whether there exist quantum simulation tasks that are suciently robust to this loss of information and can be performed by a noisy quantum simulator but are beyond the reach of classical computing is one of the primary questions being tackled in the Air Force Oce of Scientific Research (AFOSR) funded project “Experimental Robustness vs. Computational Complexity in a Neutral Atom Based NISQ Information Processor.” Whether such tasks exist in the regime of noise of current devices is both timely and important. The project is a collaborative theory-experiment effort, working to characterize where computational hardness occurs, how this correlates with the reliability of the experimental measurement for well-designed protocols, and developing a stateof- the-art neutral-atom quantum simulator platform to experimentally test, verify, and challenge our conclusions. Critical to the success of the project is the development of classical simulation methods to challenge the quantum ‘supreme’ computation performed by our quantum simulator. One of our e orts is the development of classical simulators based on representing the quantum state with tensor networks and artificial neural networks. Together, these numerical methods provide us with a multi-pronged approach to address static and dynamical properties of systems with di erent dimensionality and types of interaction. Nevertheless, addressing computational tasks of high complexity in a timely manner requires a substantial amount of computational resources, which can be met with a computing cluster with multi-core CPUs and GPUs, as described in this proposal. The computational parallelism o ered by such a setup will help facilitate the investigation of what quantum advantages are possible on noisy quantum simulators.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 20, 2023
- Source ID
- FA95502210498
Entities
People
- Tameem Albash
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of New Mexico