arXiv subjects

Jordan J. Winetrout

Publications and source records attributed to Jordan J. Winetrout.

3 recordsLinked to original sources

arXiv2026

Dataflows and Computational Patterns for Hybrid Quantum-Classical Scientific Computing

Hybrid quantum-classical computing has emerged as the dominant paradigm for near-term quantum applications, yet hybrid workflows are typically described by individual algorithms rather than their underlying execution behavior. We introduce the Quantum Execution Locality Framework (QELF), a qualitative framework for characterizing hybrid quantum-classical workflows according to recurring dataflow structures and quantum execution locality, the extent to which computation remains resident on the Quantum Processing Unit (QPU) before host intervention or classical synchronization. From a representative cross-section of applications, QELF identifies five recurring computational patterns with distinct locality characteristics and discusses their implications for communication overhead, workflow organization, and future hybrid computing architectures. By providing a common vocabulary for reasoning about hybrid workloads, QELF establishes a foundation for future quantitative validation and the co-design of algorithms, runtime systems, and hybrid computing architectures.

quant-ph↗

arXiv2024

Implementing Reactivity in Molecular Dynamics Simulations with Harmonic Force Fields

The simulation of chemical reactions and mechanical properties including failure from atoms to the micrometer scale remains a longstanding challenge in chemistry and materials science. Bottlenecks include computational feasibility, reliability, and cost. We introduce a method for reactive molecular dynamics simulations using a clean replacement of non-reactive classical harmonic bond potentials with reactive, energy-conserving Morse potentials, called the Reactive INTERFACE Force Field (IFF-R). IFF-R is compatible with force fields for organic and inorganic compounds such as IFF, CHARMM, PCFF, OPLS-AA, and AMBER. Bond dissociation is enabled by three interpretable Morse parameters per bond type and zero energy upon disconnect. Use cases for bond breaking in molecules, failure of polymers, carbon nanostructures, proteins, composite materials, and metals are shown. The simulation of bond forming reactions was included via template-based methods. IFF-R maintains the accuracy of the corresponding non-reactive force fields and is about 30 times faster than prior reactive simulation methods.

cond-mat.stat-mech↗

arXiv2021

Prediction of Carbon Nanostructure Mechanical Properties and Role of Defects Using Machine Learning

Carbon fiber and graphene-based nanostructures such as carbon nanotubes (CNTs) and defective structures have extraordinary potential as strong and lightweight materials. A longstanding bottleneck has been lack of understanding and implementation of atomic-scale engineering to harness the theoretical limits of modulus and tensile strength, of which only a fraction is routinely reached today. Here we demonstrate accurate and fast predictions of mechanical properties for CNTs and arbitrary 3D graphitic assemblies based on a training set of over 1000 stress-strain curves from cutting-edge reactive MD simulation and machine learning (ML). Several ML methods are compared and show that our newly proposed hierarchically structured graph neural networks with spatial information (HS-GNNs) achieve predictions in modulus and strength for any 3D nanostructure with only 5-10% error across a wide range of possible values. The reliability is sufficient for practical applications and a great improvement over off-the shelf ML methods with up to 50% deviation, as well as over earlier models for specific chemistry with 20% deviation. The algorithms allow more than 10 times faster mechanical property predictions than traditional molecular dynamics simulations, the identification of the role of defects and random 3D morphology, and high-throughput screening of 3D structures for enhanced mechanical properties. The algorithms can be scaled to morphologies up to 100 nm in size, expanded for chemically similar compounds, and trained to predict a broader range of properties.

cond-mat.mtrl-sci↗