SEARCH · Engineering Papers
Results for “Mathematics and Computing, Computer science”
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.
Physics-guided Machine Learning: from Supervised Deep Networks to Unsupervised Lightweight Models [Slides]
Machine learning yields great potentials in improving imaging performance (i.e., accuracy and efficiency). The incorporation of governing equation will improve generalization and alleviate label scarcity. Employ physical properties can reduce model complexity and significantly save training cost without compromising accuracy. Combining SciML imaging and edge computing would allow broader applications in energy, medicine, and other domains.
Elastic Workflows with PMIx [Slides]
Scientific workflows increasing in complexity. Elastic workflows promise improved application and system performance.
Charliecloud’s Successful Prototype Integration with Slurm: A Promising Approach with Some Strings Attached [Slides]
Abstract not provided.
Pavilion 2 Feature Additions [Slides]
Pavilion 2 is a critical component of system testing for LANL HPC systems. Pavilion 2 is system independent, which allows for the creation of a large suite of generalized tests than can be applied to multiple systems. These additions to Pavilion 2 improve our ability to test systems and diagnose issues. With such a large suite, running the proper tests when diagnosing and fixing a system can be difficult, as well as locating old tests for reference. Test configurations may contain several permutations of a test, not all of which may apply to a given system. The ability to run sets of test permutations lets the tester focus on specific parts of the test without having to edit the test configurations or run extraneous tests. The addition of the filter argument improves the ability to find test results based on test attributes, including past tests. Pavilion 2's system independence is largely based on various layers of configuration files. Configuration files for the host, test, and modes exist to allow generalized testing. The operating system configuration layer can provide helpful OS defaults to a test. Survey, a collection and reporting tool, can help us gauge the performance of a system as well as diagnose issues. The addition of a Survey mode in Pavilion 2 makes running tests with Survey very straightforward. The Survey results are combined with Pavilion 2's test results to make it easy to access.
Understanding racial and ethnic disparities in COVID-19 outcomes using an agent-based simulation [Poster]
Infectious disease outbreaks pose a major threat to public health, economic stability, and human security. In order to mitigate disease impacts, we need to understand drivers that contribute to its spread. Latinx, Black, and American Indian racial and ethnic groups experienced disproportionate health outcomes during the COVID-19 pandemic. Current modeling approaches often fail to capture the high level of heterogeneity present in the population needed to measure the drivers of these disparities.
Implement Intel Level Zero (ZE) component for Open MPI Accelerator framework [Poster]
Abstract not provided.
MTT maintenance/deployment on ECP platforms [Poster]
Abstract not provided.
LANL code team outbriefs from the El Cap Hackathon, December 2023
Goals: Build Moya with both MPI backend and Legion backend for MI300a. Accomplishments: Ristra Moya was run successfully with both MPI and Legion backend on MI300a; 2X speedup compared to MI250X was observed with MPI backend.
OpenCHAMI Introduction: Cloud-like HPC Management for the Composable future [Slides]
OpenCHAMI is: An independent partnership of large HPC operators and vendor(s); Based on the APIs and Microservices from the Cray System Manager; A set of governance standards for developing HPC system management software aligned with cloud principles; A github organization; A set of software repositories in active development to enable sites to solve their own problems in a shareable way.
PETSc/TAO Users Manual V.3.21
This manual describes the use of the Portable, Extensible Toolkit for Scientific Computation (PETSc) and the Toolkit for Advanced Optimization (TAO) for the numerical solution of partial differential equations (PDEs) and related problems on high-performance computers. PETSc/TAO is a suite of data structures and routines that provide the building blocks for implementing large-scale application codes on parallel (and serial) computers. PETSc uses the MPI standard for all distributed memory communication. PETSc/TAO includes a large suite of parallel linear solvers, nonlinear solvers, time integrators, and optimizers that may be used in application codes written in Fortran, C, C++, and Python (via petsc4py; see Getting Started ). The library is organized hierarchically, enabling users to employ the abstraction level most appropriate for a particular problem. By using techniques of object-oriented programming, PETSc provides enormous flexibility for users.
Probabilistic data fusion and physics-informed machine learning: A new paradigm for modeling under uncertainty, and its application to accelerating the discovery of new materials
In this report we summarize the work conducted by PI Perdikaris and his group under this Early Career project DE–SC0019116 during the period of 09/01/2018 – 08/31/2023. The central aim of the work was to introduce a new paradigm for scientific data analysis that can seamlessly synthesize rigorous mathematical modeling with data of variable fidelity (e.g., measurements at multiple scales/resolutions or predictions of variable fidelity models) and multiple modalities (e.g., images, time–series, or scattered measurements). The setting we are interested in involves complex systems that are partially observed and whose dynamical behavior could be hard to model or totally unknown. The inherent uncertainty associated with this setting necessitates a departure from the classical deterministic realm of modeling and scientific computation, and, consequently, our main building blocks can no longer be crisp deterministic numbers and governing laws, but instead we must operate with probabilistic models.
BB-CVXOPT: Basic Block Execution Count Estimation and Extrapolation using Constrained Convex Optimization
Explore the source record for details and available documents.
Data-aware Scheduling with the Convergence of HPC and Cloud
The overarching goal of this project was to advance the capabilities of data management and workflow optimization in molecular dynamics (MD) simulations.
Real-time Object Bounding in LiDAR Data With Computer Vision
The Multimodal Measurement System is a roadside radiation measurement testbed used to detect radiation sources in passing vehicles. It works by combining sensor signals from various modalities to produce a thorough scan of the source. A LiDAR sensor is used to measure the dimensions of the vehicle and provide a velocity estimate. However, the current LiDAR setup uses propriety software for which the source code is unavailable and cannot be updated to improve performance. Therefore, it is imperative to the accuracy of the analysis to create a custom vehicle detection that can return the dimensions and velocity of passing vehicles in real time. This new custom detection is written in C++ using the PointCloud Library, which keeps it lightweight. It also utilizes Docker and the Robot Operating System, which allows the versatility of running both on a small computer or the Lawrence Livermore National Laboratory cluster while utilizing different models of LiDAR sensors. The custom detection outperforms the current detection model, which increases the accuracy of radiation source detection.
Structure-Aware Unsupervised, Transformational Machine Learning for Drug Discovery (DTRA Basic Research Final Report)
The major goal of this project is to develop machine learning (ML) methods to enable improved predictive power on real drug discovery for novel targets. More specifically, we planned to demonstrate the capability and effectiveness of ML tools utilizing unlabeled large-volume protein-ligand datasets. We investigated multiple pre-training approaches for 3D protein-ligand structure-based foundation models, without relying on experimental binding data. We also addressed scenarios in which crystal structures are unavailable or binding data are limited. We also planned to develop a complete pipeline to screen novel compounds as well as to demonstrate the capability and effectiveness of the developed methods by testing on a realistic drug discovery task such as SARS-CoV-2. While the major goals and milestones remain consistent with the original proposal, certain technical details have been adjusted, based on the experimental results and related outcomes.
Data Center Market Report
The data center market is poised to explode in the coming decade due to undeniable drivers such as continued adoption of generative AI, increased data storage needs, and enterprise integration of AI in numerous industries [1] [2] [3]. Scalable power and increased computational capacity are at the forefront of considerations for hyperscalers, the major cloud service providers in this space. Lawrence Livermore National Laboratory is uniquely poised to help with informed decision making for data center market leaders during this phase of explosive expansion. National grid modeling expertise and cutting edge innovations in computer cooling systems place LLNL in an enviable position for creating economic impact in the data center industry by leveraging its expertise in these areas which can help the data center market keep up with growing demand.
Watching the Watchers with Verified Formal-Assurance Tools (Abbreviated Final Report)
The “Watching the Watchers” project studied the problem of establishing assurance cases for tools that are used to assure other things. Specifically, we were interested in understanding the tools and techniques one could apply to software to build an assurance case to evaluate their applicability, difficulty, level of assurance provided, and scalability. To do so we chose a set of use cases of relevance to LLNL and our various DOE and non-DOE partners and developed demonstrators to perform this evaluation. Our key focal point was around additive manufacturing problems and assurance gaps that we identified in the additive manufacturing workflow from start to completion. We also explored other areas related to AI, data analysis, and concurrent programming. Follow-on research is planned to take our prototypes from this project and adapt and mature them to fit LLNL mission applications.