Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “heuristic”

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 1,027 records · Page 57

RAP: Resource-aware Automated GPU Sharing for Multi-GPU Recommendation Model Training and Input Preprocessing

Ensuring high-quality recommendations for newly onboarded users requires the continuous retraining of Deep Learning Recommendation Models (DLRMs) with freshly generated data. To serve the online DLRM retraining, existing solutions use hundreds of CPU computing nodes designated for input preprocessing, causing significant power consumption that surpasses even the power usage of GPU trainers. To this end, we propose RAP, an end-to-end DLRM training framework that supports Resource-aware Automated GPU sharing for DLRM input Preprocessing and Training. The core idea of RAP is to accurately capture the remaining GPU computing resources during DLRM training for input preprocessing, achieving superior training efficiency without requiring additional resources. Specifically, RAP utilizes a co-running cost model to efficiently assess the costs of various input preprocessing operations, and it implements a resource-aware horizontal fusion technique that adaptively merges smaller kernels according to GPU availability, circumventing any interference with DLRM training. In addition, RAP leverages a heuristic searching algorithm that jointly optimizes both the input preprocessing graph mapping and the co-running schedule to maximize the end-to-end DLRM training throughput. The comprehensive evaluation shows that RAP achieves 78.3× speedup on average over CPU-based DLRM input preprocessing frameworks. In addition, the end-to-end training throughput of RAP is only 2.04% lower than the ideal case, which has no input preprocessing overhead.

Wang, Zheng↗

QASMTrans: A QASM Quantum Transpiler Framework for NISQ Devices

In quantum computing, transpilation plays a crucial role in converting high-level, machine-independent quantum circuits into circuits specially for a quantum device, considering factors such as basis gate set, topology, error profile, etc. Yet, the efficiency of transpilation remains a significant bottleneck, particularly when dealing with very large QASM level input files. In this paper, we present QASMTrans, a C++ based high-performance quantum transpiler framework that can demonstrate on average 50-100× speedups compared to the internal transpiler of Qiskit. Particularly, for large dense circuits such as ’uccsd n24’ and ’qft n320’ incorporating millions of gates, QASMTrans can successfully transpile in 69s and 31s, respectively, while Qiskit failed to finish in one hour. Using QASMTrans as the baseline, it becomes more feasible to explore much larger design space and impose more comprehensive compiler optimizations.

Hua, Fei↗

cuAlign: Scalable Network Alignment on GPU Accelerators

Given two graphs, the objective of network alignment is to find the best one-to-one mapping of vertices in one graph (??) to vertices in the other (??), such that the number of overlaps is maximized. We say that edges(??, ??) ???and(??', ??') ??? are overlapped if ?? is mapped to ??' and ?? is mapped to??'. Network alignment is an important optimization problem with several applications in bioinformatics, computer vision and ontology matching. Since it is an NP-hard problem, efficient heuristics and scalable implementations are necessary. In this work, we introduce a new framework that combines the concepts of intra-network proximity using vertex embedding,Belief Propagation (BP) and approximate weighted matching, and provides qualitative improvements up to22%over state-of-the-art approaches. We also provide scalable implementations on GPU accelerators, demonstrating up to19×speedup for Belief Propagation and 3× speedup for approximate weighted matching relative to previous multithreaded implementation. A combination of combinatorial and algebraic kernels within the network alignment algorithm poses significant hurdles for parallelization. Load imbalance and irregular DRAM traffic limit achievable performance on GPUs. Our novel approach identifies and exploits unique structural proper-ties of the BP-based algorithm and employs code fusion to reduce data movement between different steps of the algorithm. Using a diverse set of inputs, we demonstrate qualitative improvements of our algorithms, and performance gains of our GPU-accelerated implementation. We believe that our work will enable algorithmic improvements and practical applications of network alignment.

Xiang, Lizhi↗

Divergence Reduction in Monte Carlo Neutron Transport with On-GPU Asynchronous Scheduling

While Monte Carlo Neutron Transport (MCNT) is near-embarrasingly parallel, the effectively unpredictable lifetime of neutrons can lead to divergence when MCNT is evaluated on GPUs. Divergence is the phenomenon of adjacent threads in a warp executing different control flow paths; on GPUS, it reduces performance because each work group may only execute one path at a time. The process of Thread Data Remapping (TDR) resolves these discrepancies by moving data across hardware such that data in the same warp will be processed through similar paths. A common issue among prior implementations of TDR is the synchronous nature of its remapping and processing cycles, which exhaustively sort data produced by prior processing passes and exhaustively evaluate the sorted data. In another work, we defined a method of remapping data through an asynchronous scheduler which allows for work to be stored in shared memory and deferred arbitrarily until that work is a viable option for low-divergence evaluation. This article surveys a wider set of cases, with the goal of characterizing performance trends across a more comprehensive set of parameters. These parameters include cross sections of scattering/capturing/fission, use of implicit capture, source neutron counts, simulation time spans, and tuned memory allocations. Across these cases, we have recorded minimum and average execution times, as well as a heuristically tuned near-optimal memory allocation size for both synchronous and asynchronous scheduling. Across the collected data, it is shown that the asynchronous method is faster and more memory efficient in the majority of cases, and that it requires less tuning to achieve competitive performance.

Computer Science↗

Extending High-Level Synthesis with AI/ML Methods

Artificial Intelligence (AI) and Machine Learning (ML) methods provide significant opportunities of improving quality of results when performing high-level synthesis (HLS). For example, they can be used to model and predict metrics of the final design (e.g., area, considering aspects such as interconnect overhead for different device technologies), facilitating exploration when searching for the best design trade-offs. They can also enable identifying hidden correlations across the various phases of the synthesis and the various optimizations performed, identifying the most effective pipelines. Finally, in more general terms, bio-inspired heuristic algorithms can improve the design space exploration for the synthesis process in terms of time and quality of the result. This paper discusses opportunities and challenges to augment HLS with AI/ML using as example flow the SODA Synthesizer, an open-source hardware generation toolchain which includes SODA-OPT, a hardware/software partitioning and pre-optimization tool developed with the MLIR framework, and PandA-Bambu, a state-of-the art HLS tool. SODA interfaces with OpenROAD to provide a complete end-to-end toolchain.

artificial intelligence↗

ScaWL: Scaling k-WL (Weisfeiler-Lehman) Algorithms in Memory and Performance on Shared and Distributed-Memory Systems

The k-dimensional Weisfeiler-Lehman (k-WL) algorithm—developed as an efficient heuristic for testing if two graphs are isomorphic—is a fundamental kernel for node embedding in the emerging field of graph neural networks. Unfortunately, the k-WL algorithm has exponential storage requirements, limiting the size of graphs that can be handled. This work presents a novel k-WL scheme with a storage requirement orders of magnitude lower while maintaining the same accuracy as the original k-WL algorithm. Due to the reduced storage requirement, our scheme allows for processing much bigger graphs than previously possible on a single compute node. For even bigger graphs, we provide the first distributed-memory implementation. Our k-WL scheme also has significantly reduced communication volume and offers high scalability. Our experimental results demonstrate that our approach is significantly faster and has superior scalability compared to five other implementations employing state-of-the-art techniques.

algorithims↗

Foundation Models for Zero-Shot Segmentation of Scientific Images without AI-Ready Data

Zero-shot and prompt-based models have excelled at visual reasoning tasks by leveraging large-scale natural image corpora, but they often fail on sparse and domain-specific scientific image data. We introduce Zenesis, a no-code interactive computer vision platform designed to reduce data readiness bottlenecks in scientific imaging workflows. Zenesis integrates lightweight multimodal adaptation for zero-shot inference on raw scientific data, human-in-the-loop refinement, and heuristic-based temporal enhancement. We validate our approach on Focused Ion Beam Scanning Electron Microscopy (FIB-SEM) datasets of catalyst-loaded membranes. Zenesis outperforms baselines, achieving an average accuracy of 0.947, Intersection over Union (IoU) of 0.858, and Dice score of 0.923 on amorphous catalyst samples; and 0.987 accuracy, 0.857 IoU, and 0.923 Dice on crystalline samples. These results represent a significant performance gain over conventional methods such as Otsu thresholding and standalone models like the Segment Anything Model (SAM). Zenesis enables effective image segmentation in domains where annotated datasets are limited, offering a scalable solution for scientific discovery.

Mukherjee, Shubhabrata↗

Self-Admitted Technical Debt in Scientific Software: Prioritization, Sentiment, and Propagation Across Artifacts

Self-admitted technical debt (SATD) impairs scientific software (SSW), yet its prioritization, sentiment, persistence, and propagation remains underexplored. Understanding how SSW developers express, and address SATD is crucial for improving SSW maintenance, and tooling. This study investigates how SATD types and artifacts in SSW are prioritized, how sentiment relates to urgency, SATD removal and resolution rates, and the extent to which SATD propagates across artifacts. We analyzed nine SSW repositories using a SATD classification model and a semantic embedding-based prioritization heuristic. SATD was examined across multiple artifacts, with sentiment assessed via a fine-tuned transformer. Propagation was traced, priority scores compared to static analysis, and removal and resolution rates quantified. SATD in comments, commits, and pull requests receive higher priority than SATD in issues, with negative sentiment amplifying urgency. Resolution and removal rates lag behind open-source software (OSS) averages. Most SATD remains confined to the originating artifact, but longer propagation chains are rare and correlate with higher priority, highlighting persistent and high impact debt. Prioritization is influenced by artifact type and sentiment, while low removal and resolution rates signal persistent debt. Cross-artifact propagation marks high priority, unresolved SATD, providing empirical guidance for targeted monitoring, review prioritization, and tool supported maintenance in SSW.

Melin, Eric [Boise State University]↗

Topology Optimization of 3D Flow Fields for Flow Batteries

We report as power generated from renewables becomes more readily available, the need for power-efficient energy storage devices, such as redox flow batteries, becomes critical for successful integration of renewables into the electrical grid. An important aspect of a redox flow battery is the planar flow field, which is usually composed of two-dimensional channels etched into a backing plate. As reactant-laden electrolyte flows into the flow battery, the channels in the flow field distribute the fluid throughout the reactive porous electrode. We utilize topology optimization to design flow fields with full three-dimensional geometry variation, i.e., 3D flow fields. Specifically, we focus on vanadium redox flow batteries and use the optimization algorithm to generate 3D flow fields evolved from standard interdigitated flow fields by minimizing the electrical and flow pressure power losses. To understand how these 3D designs improve performance, we analyze the polarization of the reactant concentration and exchange current within the electrode to highlight how the designed flow fields mitigate the presence of electrode dead zones. While interdigitated flow fields can be heuristically engineered to yield high performance by tuning channel and land dimensions, such a process can be laborious; this work provides a framework for automating that design process.

25 ENERGY STORAGE↗

On Deployment of Multiple Base Stations for Energy-Efficient Communication in Wireless Sensor Networks

Data transmission from sensor nodes to a base station or a sink node often incurs significant energy consumption, which critically affects network lifetime. We generalize and solve the problem of deploying multiple base stations to maximize network lifetime in terms of two different metrics under one-hop and multihop communication models. In the one-hop communication model, the sensors far away from base stations always deplete their energy much faster than others. We propose an optimal solution and a heuristic approach based on the minimal enclosing circle algorithm to deploy a base station at the geometric center of each cluster. In the multihop communication model, both base station location and data routing mechanism need to be considered in maximizing network lifetime. We propose an iterative algorithm based on rigorous mathematical derivations and use linear programming to compute the optimal routing paths for data transmission. Simulation results show the distinguished performance of the proposed deployment algorithms in maximizing network lifetime.

Lin, Yunyue↗

Spontaneous multi-keV electron generation in a low-RF-power axisymmetric mirror machine

Title:Spontaneous multi-keV electron generation in a low-RF-power axisymmetric mirror machineAbstract:X-ray emission shows the existence of multi-keV electrons in low-temperature, low-power, capacitively-coupled RF-heated magnetic-mirror plasmas that also contain a warm (300 eV) minority electron population. Though these warm electrons are initially passing particles, we suggest that collisionless scattering -- mu non-conservation in the static vacuum field -- is responsible for a minority of them to persist in the mirror cell for thousands of transits during which time a fraction are energized to a characteristic temperature of 3 keV, with some electrons reaching energies above 30 keV. A heuristic model of the heating by a Fermi-acceleration-like mechanism is presented, with mu non-conservation in the static vacuum field as an essential feature.

Swanson, C↗

CanTrilBat and Cantera_apps v1.0 beta

This application can determine the performance and chemical behavior of batteries in 1D when they are cycled. With CanTrilBat, we are developing predictive phenomenological models for battery systems to predict operating performance and rate limiting steps in the performance of battery models. Particular attention is paid to primary and secondary chemistry mechanisms, such as the thermal runaway mechanisms experienced in secondary lithium ion batteries or self-discharge reaction mechanism that all batteries experience to one extent or another. The first application of this model has been for modeling the performance of thermal batteries. However, an implementation for secondary ion batteries is next. CanTrilBat applications solves transient problems involving batteries. It is a 1-D application that represents 3-D physical systems that can be reduced using the porous flow approximation for the anode, cathode, and separator. A control volume formulation is used to track conserved quantities. An operator-split approach is used to calculate the chemistry, diffusion and electronic transport that occurs within cathode and anode particles, allowing for the reduction in code complexity. All jacobian operations in CanTrilBat utilize numerical jacobians. A home grown predictor corrector scheme is used for time step control, and a home grown newton solver is used to relax the equations at each time step. Trilinos is used to solve for the resulting linear systems equation A block text input format is used to initialize options for the CanTriBat program. Within this block are the names of several XML input files, which specify the chemistry mechanism. These XML input files are read by low level Cantera routines, and serve to initialize the electrode and electrolyte chemistry mechanisms and transport properties. A GUI implementation has been contracted out to a university professor, but has not been implemented yet. It?s expected that CanTriBat will have the capability to do adaptive grid refinement in later versions. Grids will be halved based on an complicated set of heuristic criteria, based on how well the mesh captures changes to the first and second derivatives of selected solution components. Primary emphasis is given towards converged profiles for IV performance.

Moffat, HarryK.↗

Topology-aware task-assIGnment mappER (TIGER) v0.1

The general topology-aware task assignment problem is an NP-complete graph isomorphism problem. Existing task scheduling approaches are either heuristic or based on physical optimization algorithms, providing different speed and solution quality trade-offs. Physical Ising machines such as quantum and digital annealers have recently become available offering an alternative hardware solution to solve certain types of optimization problems. We propose an algorithm that allows expressing the problem in the compatible format. We present a domain specific partition strategy that enables to execute larger scale problems. TIGER - topology-aware task assignment mapper tool - implements the proposed algorithm and automatically integrates task- communication graph and an architecture graph into the quantum software environment.

Donofrio, David↗

Xavier: High-Performance X-Drop Adaptive Banded Pairwise Alignment (Xavier) v1.0

Pairwise sequence alignment is one of the most compute-intense challenges for the analysis of genomics data and it can account for more than 90% of the runtime depending on the application. Despite the effort to reduce the computational cost introducing heuristics such as banded and x-drop alignment, the performance of state-of-the-art software is often unsatisfactory for long sequences. Recent advances in single-molecule sequencing technologies keep increasing the length of the sequences to align at the cost of higher error rates b ranging from 5% to 15% b, emphasizing the need for fast and accurate software. Our long-read alignment software, named Xavier, computes high-performance x-drop adaptive banded pairwise alignment exploiting instruction-level parallelism via Single-Instruction-Multiple-Data (SIMD) operations. Xavier uses a narrow bandwidth that appreciably improves performance reducing the search space for the optimal alignment. In our experiment, the adaptive mechanism enables Xavier to find the optimal alignment between related sequences even when the error rate is as high as 50%. Xavier achieves speed-up up to 5X over competitor software when enabling the x-drop termination and up to 415X when disabling it, with negligible accuracy loss.

Guidi, Giulia↗

STreCH

Symbolic regression software that formulates a symbolic regression problem as a mixed-integer nonlinear optimization problem, and solves it using a sequential tree construction heuristic

Leyffer, Sven↗

MIDAS-DC-AC Tool: Fully Automating the Acquisition of AC Power Flow Solution (MIDAS-DC2AC)

MIDAS DC2AC is an automated tool for achieving a converged AC power flow solution from any dispatch, e.g., determined using DC power flow model based optimal power flow. The entire process is free of human interference. It is usually encountered in practice that, even with a DC power flow solution, acquiring the solved AC power flow solution, if exists, sometimes could be a challenging task, especially during the planning stage. It is also difficult to distinguish the unsolvable cases from diverging iterations. Manual adjustments to approach the desired power flow condition has been largely relied on in the past using lots of engineering heuristics. This tool provides a systematic way to first achieve a solvable AC power flow case by modifying the power flow condition, and then try to track the AC power flow solution while gradually removing the adopted changes. If all adopted changes can be completely removed, then the original AC power flow solution is obtained. Otherwise, insights for actionable controls are derived to help operation and planning. Currently, this tool has been implemented in Python using SIEMENS PTI PSS/E as power flow solver, where only adjusting generator terminal voltage set point is considered as an available means to try to turn an unsolved power flow to a solved one. In future, more means should be considered, including the operation of tap-changing transformers, switched shunts and redispatch of active power.

Tan, Jin↗

Code for Value Decomposition Graph Network and environment for AMR on linear advection

This is the code for the paper [Multi-Agent Reinforcement Learning for Adaptive Mesh Refinement](https://arxiv.org/abs/2211.00801), published at AAMAS 2023. It contains the implementation of a new algorithm, called Value Decomposition Graph Network (VDGN), for applying multi-agent reinforcement learning to the problem of adaptive mesh refinement (AMR). It also contains the implementation of a multi-agent environment for AMR on a linear advection problem. VDGN is the first learning algorithm to display anticipatory refinement behavior in AMR, and it outperforms local error threshold-based heuristic strategies.

Yang, Jiachen↗

PV O&M Cost Model [SWR-21-14]

This tool presents a method for calculating costs associated with the operation and maintenance (O&M) of photovoltaic (PV) systems. The tool compiles details regarding the cost and frequency of multiple O&M services to estimate annual O&M costs ($/year) for each year of an analysis period, the net present value ($) of life cycle costs accumulated over the analysis period, and the reserve account amount ($) that might be required to fund unexpected repairs. The reserve account includes parts inventory and is important for providing a source of funds to make repairs quickly and avoid lost production. This method allows a detailed selection of services to perform based on system size, market served (e.g., residential, commercial, or utility), type and configuration of system components (e.g., micro-, string, or central inverter), and site and environmental conditions (e.g., pollen, bird populations) which is an improvement over simple per unit valuations of O&M costs ($/kW/year). This model also distinguishes costs that vary from year to year and increase at different rates over time as modeled by heuristic failure distributions (e.g., Weibull or Lognormal distribution) based on actuarial data for many of the services.

Walker, Andy↗