Quantum Digital Twins for Uncertainty Quantification

TL;DR

This paper introduces quantum digital twins as virtual models of quantum processors to analyze noise and emulate parallel quantum computing, aiming to accelerate quantum advantage in resource-intensive tasks.

Contribution

The paper develops quantum digital twins to simulate and analyze quantum noise and enable distributed quantum computing, advancing practical quantum applications.

Findings

Quantum digital twins effectively model quantum noise.

Hybrid quantum ensembles improve distributed quantum processing.

Quantum digital twins help achieve early quantum advantage.

Abstract

Modern supercomputers can handle resource-intensive computational and data-driven problems in various industries and academic fields. These supercomputers are primarily made up of traditional classical resources comprising CPUs and GPUs. Integrating quantum processing units with supercomputers offers the potential to accelerate and manage computationally intensive subroutines currently handled by CPUs or GPUs. However, the presence of noise in quantum processing units limits their ability to provide a clear quantum advantage over conventional classical resources. Hence, we develop and construct "quantum digital twins," virtual versions of quantum processing units. To demonstrate the potential benefit of quantum digital twins, we create and deploy hybrid quantum ensembles on five quantum digital twins that emulate parallel quantum computers since hybrid quantum ensembles are suitable for distributed computing. Our study demonstrates that quantum digital twins assist in analyzing the actual quantum device noise on real-world use cases and emulate parallel quantum processing units for distributed computational tasks to obtain quantum advantage as early as possible.