Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “computational complexity”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,387 records · Page 77

Growth interactions between icosahedral quasicrystals

We investigate the solidification pathways of a population of icosahedral quasicrystals in a liquid through in situ synchrotron x-ray tomography. The wealth of three-dimensional space- and time-resolved data enables us to test the predictions of various models and theories of crystallization on a quasi-crystal-forming alloy. Remarkably, we find the general evolution equation—that accounts for the competing effects of growth, coalescence, and coarsening—fits well our experimental data on surface area concentration. Furthermore, we quantify the orientation selection, screening length, and coarsening rate of the dodecahedra, and compare these results to that of periodic crystals. The latter is a full order of magnitude smaller than that of elemental metal dendrites in the limit of zero volume fraction, a reflection of the low solid-liquid interfacial energy of the icosahedral phase. Furthermore, our paper provides the critical input data for microstructural models used for integrated computational materials engineering of complex intermetallics, including quasicrystals.

36 MATERIALS SCIENCE↗

A Real-Time Implementation and Validation of Federated Learning for Grid Services

Grid-edge devices are becoming increasingly important in the energy transition. Preserving privacy was not previously considered an important aspect for power grid operations, but with the increased proliferation of customer-owned assets, it is now an essential consideration. Several mechanisms have been proposed to provide privacy for non-utility owned assets in the power grid. Federated learning (FL) is one method gaining prominence in this area. Although FL has been used for other applications, such as auto-complete in phones, there has not been much investigation into whether these approaches are feasible for grid applications. In this work, we use a research platform with real-time simulators and hardware-in-the-loop capabilities to investigate how FL can be applied to grid-edge devices, and we present the potential grid services that can be derived for these devices. We discuss the computational challenges with deploying complex FL approaches, and we explore several grid services, including participation in retail electricity markets, voltage control, and resilience-driven reconfiguration.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Distilling Knowledge from Ensembles of Cluster-Constrained-Attention Multiple-Instance Learners for Whole Slide Image Classification

The peculiar nature of whole slide imaging (WSI), digitizing conventional glass slides to obtain multiple high resolution images which capture microscopic details of a patient’s histopathological features, has garnered increased interest from the computer vision research community over the last two decades. Given the unique computational space and time complexity inherent to gigapixel-size whole slide image data, researchers have proposed novel machine learning algorithms to aid in the performance of diagnostic tasks in clinical pathology. One effective algorithm represents a Whole slide image as a bag of smaller image patches, which can be represented as low-dimension image patch embeddings. Weakly supervised deep-learning methods, such as cluster-constrained-attention multiple instance learning (CLAM), have shown promising results when combined with image patch embeddings. While traditional ensemble classifiers yield improved task performance, such methods come with a steep cost in model complexity. Through knowledge distillation, it is possible to retain some performance improvements from an ensemble, while minimizing costs to model complexity. In this work, we implement a weakly supervised ensemble using clustering-constrained-attention multiple-instance learners (CLAM), which uses attention and instance-level clustering to identify task salient regions and feature extraction in whole slides. By applying logit-based and attention-based knowledge distillation, we show it is possible to retain some performance improvements resulting from the ensemble at zero cost to model complexity.

Alamudun, Folami↗

Efficient Unitary Designs from Random Sums and Permutations

A unitary k-design is an ensemble of unitaries that matches the first k moments of the Haar measure. In this work, we provide two efficient constructions of k-designs on n-qubits using new random matrix theory techniques. Our first construction is based on exponentiating sums of random i.i.d. Hermitian matrices and uses O(k2n2)-many gates. In the spirit of central limit theorems, we show that this random sum approximates the Gaussian Unitary Ensemble (GUE). We then show that the product of just two exponentiated GUE matrices is already approximately Haar random. Our second construction is based on products of exponentiated sums of random permutations and uses Õ(k poly (n)) many gates. The k dependence is optimal (up to polylogarithmic factors) and is inherited from the efficiency of existing k-wise independent permutations. Furthermore, replacing random permutations with quantum-secure pseudorandom permutations (PRPs), we also obtain a pseudorandom unitary (PRU) ensemble that is secure under nonadaptive queries. A central feature of both proofs is a new connection between the polynomial method in quantum query complexity and the large-dimension (N) expansion in random matrix theory. In particular, the first construction uses the polynomial method to control high moments of certain random matrix ensembles without requiring delicate Weingarten calculations. In doing so, we define and solve a moment problem on the unit circle, asking whether a finite number of equally weighted points can reproduce a given set of moments. In our second construction, the key step is to exhibit an orthonormal basis for irreducible representations of the partition algebra that has a low-degree large-N expansion. This allows us to show that the distinguishing probability is a low-degree rational polynomial of the dimension N.

algebra↗

On the Impact of Widening Vector Registers on Sequence Alignment

Vector extensions, such as SSE, have been part of the x86 since the 1990s, with applications in graphics, signal processing, and scientific applications. Although many algorithms and applications can naturally benefit from automatic vectorization techniques, there are still many that are difficult to vectorize due to their dependence on irregular data structures, dense branch operations, or data dependencies. Sequence alignment, one of the most widely used operations in bioinformatics workflows, has a computational footprint that features complex data dependencies. In this paper, we demonstrate that the trend of widening vector registers adversely affects the state-of-the-art sequence alignment algorithm based on striped data layouts. We present a practically efficient SIMD implementation of a parallel scan based sequence alignment algorithm that can better exploit wider SIMD units. We conduct comprehensive workload and use case analyses to characterize the relative behavior of the striped and scan approaches and identify the best choice of algorithm based on input length and SIMD width.

Needleman-Wunsch↗

On Residual Stresses and Reference Temperatures in Thermomechanical Simulations of Photovoltaic Modules Using the Finite Element Method

Thermomechanical simulation of photovoltaic (PV) modules using the finite element method (FEM) is a useful tool to evaluate module design features in terms of structural integrity, reliability, and durability. One of the main challenges in the numerical modeling of a PV module is the incorporation of residual stresses induced by the manufacturing process. Modeling assumptions and abstractions are necessary to limit the model complexity and reduce the computational time. However, oversimplifications and incorrect assumptions can lead to erroneous numerical results. Unfortunately, much simulation work still neglects process-induced stresses. This can lead to incorrect predictions of the stress-strain history and erroneous conclusions during the design process. In this work, we review current modeling practices for incorporating process-induced stresses, and contrast numerical models that consider residual stresses with those that neglect them. We find that the simulation objective and available material properties dictate which process steps need to be modeled, and explore in depth the modeling of residual stresses induced by the lamination process. We demonstrate that a simplified cooldown procedure at the beginning of the simulation can increase the model accuracy and discuss appropriate choices for starting and reference temperatures in the finite element model.

14 SOLAR ENERGY↗

Shrink or Substitute: Handling Process Failures in HPC Systems Using In-Situ Recovery

Efficient utilization of today's high-performance computing (HPC) systems with complex software and hardware components requires that the HPC applications are designed to tolerate process failures at runtime. With low mean-time-to-failure (MTTF) of current and future HPC systems, long running simulations on these systems requires capabilities for gracefully handling process failures by the applications themselves. In this paper, we explore the use of fault tolerance extensions to Message Passing Interface (MPI) called user-level failure mitigation (ULFM) for handling process failures without the need to discard the progress made by the application. We explore two alternative recovery strategies, which use ULFM along with application-driven in-memory checkpointing. In the first case, the application is recovered with only the surviving processes, and in the second case, spares are used to replace the failed processes, such that the original configuration of the application is restored. Our experimental results demonstrate that graceful degradation is a viable alternative for recovery in environments where spares may not be available.

Ashraf, Rizwan↗

Early Evaluation of Directive-Based GPU Programming Models for Productive Exascale Computing

Graphics Processing Unit (GPU)-based parallel computer architectures have shown increased popularity as a building block for high performance computing, and possibly for future Exascale computing. However, their programming complexity remains as a major hurdle for their widespread adoption. To provide better abstractions for programming GPU architectures, researchers and vendors have proposed several directive-based GPU programming models. These directive-based models provide different levels of abstraction, and required different levels of programming effort to port and optimize applications. Understanding these differences among these new models provides valuable insights on their applicability and performance potential. In this paper, we evaluate existing directive-based models by porting thirteen application kernels from various scientific domains to use CUDA GPUs, which, in turn, allows us to identify important issues in the functionality, scalability, tunability, and debuggability of the existing models. Our evaluation shows that directive-based models can achieve reasonable performance, compared to hand-written GPU codes.

Lee, Seyong↗

LLM-Inference-Bench: Inference Benchmarking of Large Language Models on AI Accelerators

Large Language Models (LLMs) have propelled groundbreaking advancements across several domains and are commonly used for text generation applications. However, the computational demands of these complex models pose significant challenges, requiring efficient hardware acceleration. Benchmarking the performance of LLMs across diverse hardware platforms is crucial to understanding their scalability and throughput characteristics. We introduce LLM-Inference-Bench, a comprehensive benchmarking suite to evaluate the hardware inference performance of LLMs. We thoroughly analyze diverse hardware platforms, including GPUs from Nvidia and AMD and specialized AI accelerators, Intel Habana and SambaNova. Our evaluation includes several LLM inference frameworks and models from LLaMA, Mistral, and Qwen families with 7B and 70B parameters. Our benchmarking results reveal the strengths and limitations of various models, hardware platforms, and inference frameworks. We provide an interactive dashboard to help identify configurations for optimal performance for a given hardware platform.

Chitty-Venkata, Krishna Teja↗

Enhancing Distribution System Resilience: A First-Order Meta-RL Algorithm for Critical Load Restoration

The increasing frequency of extreme events and the integration of distributed energy resources (DERs) into modern grids have elevated the need for resilient and efficient critical load restoration strategies in distribution systems. However, the stochastic nature of renewable DERs, limited energy resource availability and the intricate nonlinearities inherent in complex grid control problem make the problem challenging. Although reinforcement learning (RL) and warm-start RL methods have shown promising results, their performance often falls short in rapidly adapting to new, unseen situations and typically requires exhaustive problem-specific tuning. To address these gaps, we propose a First-Order Meta-based RL (FOM-RL) algorithm within an online framework for adaptive and robust critical load restoration. By harnessing local DERs as the enabling technology, FOM-RL allows the RL agent to swiftly adapt to new unseen scenarios by leveraging previously acquired knowledge of different tasks. Experimental results provide evidence that proposed algorithm learns more efficiently and showcases generalization capabilities across diverse set of operational scenarios. Moreover, a rigorous theoretical analysis yields a tight sublinear regret bound, sensitive to temporal variability, with a task-averaged optimality gap bounded by O(VM+D*/(Tsquare root(M))). These results suggest that optimality improves with task similarity and an increased number of tasks M, reaffirming the efficacy and scalability of the proposed approach in addressing the complexities of critical load restoration in distribution systems.

complexity theory↗

Coupling Deterministic and Monte Carlo Transport Methods for the Simulation of Gamma-Ray Spectroscopy Scenarios

Radiation transport modeling methods used in the radiation detection community fall into one of two broad categories: stochastic (Monte Carlo) and deterministic. Monte Carlo methods are typically the tool of choice for simulating gamma-ray spectrometers operating in homeland and national security settings (e.g. portal monitoring of vehicles or isotope identification using handheld devices), but deterministic codes that discretize the linear Boltzmann transport equation in space, angle, and energy offer potential advantages in computational efficiency for many complex radiation detection problems. This paper describes the development of a scenario simulation framework based on deterministic algorithms. Key challenges include: formulating methods to automatically define an energy group structure that can support modeling of gamma-ray spectrometers ranging from low to high resolution; combining deterministic transport algorithms (e.g. ray-tracing and discrete ordinates) to mitigate ray effects for a wide range of problem types; and developing efficient and accurate methods to calculate gamma-ray spectrometer response functions from the deterministic angular flux solutions. The software framework aimed at addressing these challenges is described and results from test problems that compare coupled deterministic-Monte Carlo methods and purely Monte Carlo approaches are provided.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Make the Fastest Faster: Importance Mask Synthesis for Interactive Volume Visualization using Reconstruction Neural Networks

Visualizing a large-scale volumetric dataset with high resolution is challenging due to the substantial computational time and space complexity. Recent deep learning-based image inpainting methods significantly improve rendering latency by reconstructing a high-resolution image for visualization in constant time on GPU from a partially rendered image where only a portion of pixels go through the expensive rendering pipeline. However, existing solutions need to render every pixel of either a predefined regular sampling pattern or an irregular sample pattern predicted from a low-resolution image rendering. Both methods require a significant amount of expensive pixel-level rendering. In this work, we provide Importance Mask Learning (IML) and Synthesis (IMS) networks, which are the first attempts to directly synthesize important regions of the regular sampling pattern from the user’s view parameters, to further minimize the number of pixels to render by jointly considering the dataset, user behavior, and the downstream reconstruction neural network. Our solution is a unified framework to handle various types of inpainting methods through the proposed differentiable compaction/decompaction layers. Experiments show our method can further improve the overall rendering latency of state-of-the-art volume visualization methods using reconstruction neural network for free when rendering scientific volumetric datasets. Our method can also directly optimize the off-the-shelf pre-trained reconstruction neural networks without elongated retraining.

Large-scale data↗

Fast ignition of an inertial fusion target with a solid noncryogenic fuel by an ion beam

The burning efficiency of a preliminarily compressed inertial confinement fusion (ICF) target with a solid noncryogenic fuel (deuterium-tritium beryllium hydride) upon fast central ignition by a fast ion beam is studied. The main aim of the study was to determine the extent to which the spatial temperature distribution formed under the heating of an ICF target by ion beams with different particle energy spectra affects the thermonuclear gain. The study is based on a complex numerical modeling including computer simulations of (i) the heating of a compressed target with a spatially nonuniform density and temperature distributions by a fast ion beam and (ii) the burning of the target with the initial spatial density distribution formed at the instant of maximum compression of the target and the initial spatial temperature distribution formed as a result of heating of the compressed target by the ion beam. The threshold energy of the igniting ion beam and the dependence of the thermonuclear gain on the energy deposited in the target are determined.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Optimization Methods for Large-Scale Machine Learning

Here this paper provides a review and commentary on the past, present, and future of numerical optimization algorithms in the context of machine learning applications. Through case studies on text classification and the training of deep neural networks, we discuss how optimization problems arise in machine learning and what makes them challenging. A major theme of our study is that large-scale machine learning represents a distinctive setting in which the stochastic gradient (SG) method has traditionally played a central role while conventional gradient-based nonlinear optimization techniques typically falter. Based on this viewpoint, we present a comprehensive theory of a straightforward, yet versatile SG algorithm, discuss its practical behavior, and highlight opportunities for designing algorithms with improved performance. This leads to a discussion about the next generation of optimization methods for large-scale machine learning, including an investigation of two main streams of research on techniques that diminish noise in the stochastic directions and methods that make use of second-order derivative approximations.

97 MATHEMATICS AND COMPUTING↗

Efficient and Extensible Quasi-Explicit Modular Nonlinear Multiscale Battery Model: GH-MSMD

Complex physics and long computation time hinder the adoption of computer aided engineering models in the design of large-format battery cells and systems. A modular, efficient battery simulation model -- the multiscale multidomain (MSMD) model -- was previously introduced to aid the scale-up of Li-ion material and electrode designs to complete cell and pack designs, capturing electrochemical interplay with 3-D electronic current pathways and thermal response. Here, this paper enhances the computational efficiency of the MSMD model using a separation of time-scales principle to decompose model field variables. The decomposition provides a quasi-explicit linkage between the multiple length-scale domains and thus reduces time-consuming nested iteration when solving model equations across multiple domains. In addition to particle-, electrode- and cell-length scales treated in the previous work, the present formulation extends to bus bar- and multi-cell module-length scales. We provide example simulations for several variants of GH electrode-domain models.

30 DIRECT ENERGY CONVERSION↗

Pairwise Sequence Alignment Library

Vector extensions, such as SSE, have been part of the x86 CPU since the 1990s, with applications in graphics, signal processing, and scientific applications. Although many algorithms and applications can naturally benefit from automatic vectorization techniques, there are still many that are difficult to vectorize due to their dependence on irregular data structures, dense branch operations, or data dependencies. Sequence alignment, one of the most widely used operations in bioinformatics workflows, has a computational footprint that features complex data dependencies. The trend of widening vector registers adversely affects the state-of-the-art sequence alignment algorithm based on striped data layouts. Therefore, a novel SIMD implementation of a parallel scan-based sequence alignment algorithm that can better exploit wider SIMD units was implemented as part of the Parallel Sequence Alignment Library (parasail). Parasail features: Reference implementations of all known vectorized sequence alignment approaches. Implementations of Smith Waterman (SW), semi-global (SG), and Needleman Wunsch (NW) sequence alignment algorithms. Implementations across all modern CPU instruction sets including AVX2 and KNC. Language interfaces for C/C++ and Python.

Jeff Daily, PNNL↗

Data-Driven Transition Path Analysis Yields a Statistical Understanding of Sudden Stratospheric Warming Events in an Idealized Model

Abstract Atmospheric regime transitions are highly impactful as drivers of extreme weather events, but pose two formidable modeling challenges: predicting the next event (weather forecasting) and characterizing the statistics of events of a given severity (the risk climatology). Each event has a different duration and spatial structure, making it hard to define an objective “average event.” We argue here that transition path theory (TPT), a stochastic process framework, is an appropriate tool for the task. We demonstrate TPT’s capacities on a wave–mean flow model of sudden stratospheric warmings (SSWs) developed by Holton and Mass, which is idealized enough for transparent TPT analysis but complex enough to demonstrate computational scalability. Whereas a recent article (Finkel et al. 2021) studied near-term SSW predictability, the present article uses TPT to link predictability to long-term SSW frequency. This requires not only forecasting forward in time from an initial condition, but also backward in time to assess the probability of the initial conditions themselves. TPT enables one to condition the dynamics on the regime transition occurring, and thus visualize its physical drivers with a vector field called the reactive current . The reactive current shows that before an SSW, dissipation and stochastic forcing drive a slow decay of vortex strength at lower altitudes. The response of upper-level winds is late and sudden, occurring only after the transition is almost complete from a probabilistic point of view. This case study demonstrates that TPT quantities, visualized in a space of physically meaningful variables, can help one understand the dynamics of regime transitions.

Meteorology & Atmospheric Sciences↗

The Observation of Highly Ordered Domains in Membranes with Cholesterol

Rafts, or functional domains, are transient nano- or mesoscopic structures in the exoplasmic leaflet of the plasma membrane, and are thought to be essential for many cellular processes. Using neutron diffraction and computer modelling, we present evidence for the existence of highly ordered lipid domains in the cholesterol-rich (32.5 mol%) liquid-ordered (lo) phase of dipalmitoylphosphatidylcholine membranes. The liquid ordered phase in one-component lipid membranes has previously been thought to be a homogeneous phase. The presence of highly ordered lipid domains embedded in a disordered lipid matrix implies non-uniform distribution of cholesterol between the two phases. The experimental results are in excellent agreement with recent computer simulations of DPPC/cholesterol complexes [Meinhardt, Vink and Schmid (2013). Proc Natl Acad Sci USA 110(12): 4476 4481], which reported the existence of nanometer size lo domains in a liquid disordered lipid environment.

Armstrong, Clare L↗