Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Math library”

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.

31 records · Page 2

A graphics processing unit accelerated sparse direct solver and preconditioner with block low rank compression

We present the GPU implementation efforts and challenges of the sparse solver package STRUMPACK. The code is made publicly available on github with a permissive BSD license. STRUMPACK implements an approximate multifrontal solver, a sparse LU factorization which makes use of compression methods to accelerate time to solution and reduce memory usage. Multiple compression schemes based on rank-structured and hierarchical matrix approximations are supported, including hierarchically semi-separable, hierarchically off-diagonal butterfly, and block low rank. Here, in this paper, we present the GPU implementation of the block low rank (BLR) compression method within a multifrontal solver. Our GPU implementation relies on highly optimized vendor libraries such as cuBLAS and cuSOLVER for NVIDIA GPUs, rocBLAS and rocSOLVER for AMD GPUs and the Intel oneAPI Math Kernel Library (oneMKL) for Intel GPUs. Additionally, we rely on external open source libraries such as SLATE (Software for Linear Algebra Targeting Exascale), MAGMA (Matrix Algebra on GPU and Multi-core Architectures), and KBLAS (KAUST BLAS). SLATE is used as a GPU-capable ScaLAPACK replacement. From MAGMA we use variable sized batched dense linear algebra operations such as GEMM, TRSM and LU with partial pivoting. KBLAS provides efficient (batched) low rank matrix compression for NVIDIA GPUs using an adaptive randomized sampling scheme. The resulting sparse solver and preconditioner runs on NVIDIA, AMD and Intel GPUs. Interfaces are available from PETSc, Trilinos and MFEM, or the solver can be used directly in user code. We report results for a range of benchmark applications, using the Perlmutter system from NERSC, Frontier from ORNL, and Aurora from ALCF. For a high frequency wave equation on a regular mesh, using 32 Perlmutter compute nodes, the factorization phase of the exact GPU solver is about 6.5× faster compared to the CPU-only solver. The BLR-enabled GPU solver is about 13.8× faster than the CPU exact solver. For a collection of SuiteSparse matrices, the STRUMPACK exact factorization on a single GPU is on average 1.9× faster than NVIDIA’s cuDSS solver.

97 MATHEMATICS AND COMPUTING↗

Tiling Framework for Heterogeneous Computing of Matrix based Tiled Algorithms

Tiling matrix operations can improve the load balancing and performance of applications on heterogeneous computing resources. Writing a tile-based algorithm for each operation with a traditional, hand-tuned tiling approach that uses for loops in C/C++ is cumbersome and error prone. Moreover, it must enable and support the heterogeneous memory management of data objects and also explore architecture-supported, native, tiled-data transfer APIs instead of copying the tiled data to continuous memory before the data transfer. The tiling framework provides a tiled data structure for heterogeneous memory mapping and parameterization to a heterogeneous task specification API. We have integrated our tiled framework into MatRIS (Math kernels library using IRIS). IRIS is a heterogeneous run-time framework with a heterogeneous programming model, memory model, and task execution model. Experiments reveal that the tiled framework for BLAS operations has improved the programmability of tiled BLAS and improved performance by ~20% when compared against the traditional method that copies the data to continuous memory locations for heterogeneous computing.

Miniskar, Narasinga Rao↗

IRIS-MASH: Efficient Multi-device Asynchronous Multi-Stream Heterogeneous Computing

In the rapidly evolving field of high-performance computing (HPC), effectively leveraging heterogeneous devices through asynchronous task programming is paramount. This paper presents a robust asynchronous task programming model tailored for a multi-device, multi-stream execution environment that incorporates a diverse array of heterogeneous computing units, including GPUs from various vendors and other accelerators. Current state-of-the-art task programming models provide methodologies to support asynchronous task executions, but they typically handle homogeneous devices using native programming languages, while support for heterogeneous devices is limited to frameworks like OpenCL. This gap presents significant challenges in abstracting heterogeneous devices to harness their true asynchronous capabilities effectively using their native programming languages. By implementing asynchronous task execution, our model significantly boosts the performance of tiled algorithm task graphs through overlapping data transfers with computation and enabling the simultaneous execution of multiple kernels. We integrate this approach into a heterogeneous Intelligent Runtime System (IRIS) and assess its performance using a suite of tiled algorithm benchmarks from the heterogeneous math kernels library (MatRIS) based on IRIS. Experimental results demonstrate a performance improvement ranging from 1.6 × to 2 × over IRIS without asynchronous support, and a notable 22% performance enhancement compared to established runtime systems such as StarPU and PaRSEC. This approach significantly improves computation efficiency of HPC workflows and provides a solid base for future exploration and development in the area of asynchronous task programming in heterogeneous systems.

Miniskar, Narasinga Rao [ORNL] (ORCID:000000018259↗

Project LASER

NASA formally launched Project LASER (Learning About Science, Engineering and Research) in March 1990, a program designed to help teachers improve science and mathematics education and to provide 'hands on' experiences. It featured the first LASER Mobile Teacher Resource Center (MTRC), is designed to reach educators all over the nation. NASA hopes to operate several MTRCs with funds provided by private industry. The mobile unit is a 22-ton tractor-trailer stocked with NASA educational publications and outfitted with six work stations. Each work station, which can accommodate two teachers at a time, has a computer providing access to NASA Spacelink. Each also has video recorders and photocopy/photographic equipment for the teacher's use. MTRC is only one of the five major elements within LASER. The others are: a Space Technology Course, to promote integration of space science studies with traditional courses; the Volunteer Databank, in which NASA employees are encouraged to volunteer as tutors, instructors, etc; Mobile Discovery Laboratories that will carry simple laboratory equipment and computers to provide hands-on activities for students and demonstrations of classroom activities for teachers; and the Public Library Science Program which will present library based science and math programs.

Source record↗

NASA Science4Girls and Their Families: Connecting Local Libraries with NASA Scientists and Education Programs to Engage Girls in STEM

NASA Science4Girls and Their Families (NS4G) partners NASA Science Mission Directorate (SMD) education programs with public libraries to provide hands-on science, technology, engineering, and math (STEM) activities and career information for girls and their families, along with training for librarians, in conjunction with Women's History Month (March). NS4G is a collaboration among education teams within the four NASA SMD education and public outreach (E/PO) Forums: Planetary, Earth, Astrophysics, and Heliophysics. It began in 2012 as an Astrophysics-led program (Astro4Girls) with 9 events around the country. Upon expanding among the four Forums, over 73 events were held in Spring 2013 (Fig. 1), with preparations underway for events in Spring 2014. All events are individually evaluated by both the student participants and participating librarians to assess their effectiveness in addressing audience needs.

SCIENCE4GIRLS↗

A Fixed Point VHDL Component Library for a High Efficiency Reconfigurable Radio Design Methodology

Advances in Field Programmable Gate Array (FPGA) technologies enable the implementation of reconfigurable radio systems for both ground and space applications. The development of such systems challenges the current design paradigms and requires more robust design techniques to meet the increased system complexity. Among these techniques is the development of component libraries to reduce design cycle time and to improve design verification, consequently increasing the overall efficiency of the project development process while increasing design success rates and reducing engineering costs. This paper describes the reconfigurable radio component library developed at the Software Defined Radio Applications Research Center (SARC) at Goddard Space Flight Center (GSFC) Microwave and Communications Branch (Code 567). The library is a set of fixed-point VHDL components that link the Digital Signal Processing (DSP) simulation environment with the FPGA design tools. This provides a direct synthesis path based on the latest developments of the VHDL tools as proposed by the BEE VBDL 2004 which allows for the simulation and synthesis of fixed-point math operations while maintaining bit and cycle accuracy. The VHDL Fixed Point Reconfigurable Radio Component library does not require the use of the FPGA vendor specific automatic component generators and provide a generic path from high level DSP simulations implemented in Mathworks Simulink to any FPGA device. The access to the component synthesizable, source code provides full design verification capability:

Hoy, Scott D.↗

Generic Discretization Library

The GenDiL library is a collection of C++ software abstractions designed to discretize and solve partial differential equations (PDEs) for high-performance computing (HPC) applications. Its primary focus is on modern C++ generic programming, which helps ensure portability across various hardware architectures. The central idea behind the library is to provide building blocks for numerical algorithms-such as discretization methods and iteration patterns-so that domain experts can focus on the math, rather than the low-level details of hardware or implementation. By defining abstractions for data types, iteration over computational grids, and scheduling of operations, the library isolates the high-level PDE algorithms from the platform-specific optimizations needed to achieve efficient performance.

Dudouit, Yohann [Lawrence Livermore National Labor↗

Software Helps Retrieve Information Relevant to the User

The Adaptive Indexing and Retrieval Agent (ARNIE) is a code library, designed to be used by an application program, that assists human users in retrieving desired information in a hypertext setting. Using ARNIE, the program implements a computational model for interactively learning what information each human user considers relevant in context. The model, called a "relevance network," incrementally adapts retrieved information to users individual profiles on the basis of feedback from the users regarding specific queries. The model also generalizes such knowledge for subsequent derivation of relevant references for similar queries and profiles, thereby, assisting users in filtering information by relevance. ARNIE thus enables users to categorize and share information of interest in various contexts. ARNIE encodes the relevance and structure of information in a neural network dynamically configured with a genetic algorithm. ARNIE maintains an internal database, wherein it saves associations, and from which it returns associated items in response to a query. A C++ compiler for a platform on which ARNIE will be utilized is necessary for creating the ARNIE library but is not necessary for the execution of the software.

Mathe, Natalie↗

Math Description Engine Software Development Kit

The Math Description Engine Software Development Kit (MDE SDK) can be used by software developers to make computer-rendered graphs more accessible to blind and visually-impaired users. The MDE SDK generates alternative graph descriptions in two forms: textual descriptions and non-verbal sound renderings, or sonification. It also enables display of an animated trace of a graph sonification on a visual graph component, with color and line-thickness options for users having low vision or color-related impairments. A set of accessible graphical user interface widgets is provided for operation by end users and for control of accessible graph displays. Version 1.0 of the MDE SDK generates text descriptions for 2D graphs commonly seen in math and science curriculum (and practice). The mathematically rich text descriptions can also serve as a virtual math and science assistant for blind and sighted users, making graphs more accessible for everyone. The MDE SDK has a simple application programming interface (API) that makes it easy for programmers and Web-site developers to make graphs accessible with just a few lines of code. The source code is written in Java for cross-platform compatibility and to take advantage of Java s built-in support for building accessible software application interfaces. Compiled-library and NASA Open Source versions are available with API documentation and Programmer s Guide at http:/ / prim e.jsc.n asa. gov.

Shelton, Robert O.↗

Alternative regularizations for Outer-Approximation algorithms for convex MINLP

In this work, we extend the regularization framework from Kronqvist et al. (Math Program 180(1):285–310, 2020) by incorporating several new regularization functions and develop a regularized single-tree search method for solving convex mixed-integer nonlinear programming (MINLP) problems. We propose a set of regularization functions based on distance metrics and Lagrangean approximations, used in the projection problem for finding new integer combinations to be used within the Outer-Approximation (OA) method. The new approach, called Regularized Outer-Approximation (ROA), has been implemented as part of the open-source Mixed-integer nonlinear decomposition toolbox for Pyomo—MindtPy. We compare the OA method with seven regularization function alternatives for ROA. Moreover, we extend the LP/NLP Branch and Bound method proposed by Quesada and Grossmann (Comput Chem Eng 16(10–11):937–947, 1992) to include regularization in an algorithm denoted RLP/NLP. We provide convergence guarantees for both ROA and RLP/NLP. Finally, we perform an extensive computational experiment considering all convex MINLP problems in the benchmark library MINLPLib. The computational results show clear advantages of using regularization combined with the OA method.

Convex Mixed-integer nonlinear programming↗

Intrinsic alignment demographics for next-generation lensing: Revealing galaxy property trends with DESI Y1 direct measurements

We present direct measurements of the intrinsic alignments (IA) of over 2 million spectroscopic galaxies using DESI Data Release 1 and imaging from four lensing surveys: DES, HSC, KiDS, and SDSS. In this uniquely data-rich regime, we take initial steps towards a more tailored IA modelling approach by building a library of IA measurements across colour, luminosity, stellar mass, and redshift. We map the dependence between galaxy type -- in terms of rest-frame colour, strength of the 4000 Angstrom break, and specific star formation rate -- and IA amplitude; the bluest galaxies have an alignment consistent with zero, across low ($0.05

Siegel, J. [Princeton U.]↗

Modeling and Simulation of Variable Mass, Flexible Structures

The advent of the new Ares I launch vehicle has highlighted the need for advanced dynamic analysis tools for variable mass, flexible structures. This system is composed of interconnected flexible stages or components undergoing rapid mass depletion through the consumption of solid or liquid propellant. In addition to large rigid body configuration changes, the system simultaneously experiences elastic deformations. In most applications, the elastic deformations are compatible with linear strain-displacement relationships and are typically modeled using the assumed modes technique. The deformation of the system is approximated through the linear combination of the products of spatial shape functions and generalized time coordinates. Spatial shape functions are traditionally composed of normal mode shapes of the system or even constraint modes and static deformations derived from finite element models of the system. Equations of motion for systems undergoing coupled large rigid body motion and elastic deformation have previously been derived through a number of techniques [1]. However, in these derivations, the mode shapes or spatial shape functions of the system components were considered constant. But with the Ares I vehicle, the structural characteristics of the system are changing with the mass of the system. Previous approaches to solving this problem involve periodic updates to the spatial shape functions or interpolation between shape functions based on system mass or elapsed mission time. These solutions often introduce misleading or even unstable numerical transients into the system. Plus, interpolation on a shape function is not intuitive. This paper presents an approach in which the shape functions are held constant and operate on the changing mass and stiffness matrices of the vehicle components. Each vehicle stage or component finite element model is broken into dry structure and propellant models. A library of propellant models is used to describe the distribution of mass in the fuel tank or Solid Rocket Booster (SRB) case for various propellant levels. Based on the mass consumed by the liquid engine or SRB, the appropriate propellant model is coupled with the dry structure model for the stage. Then using vehicle configuration data, the integrated vehicle model is assembled and operated on by the constant system shape functions. The system mode shapes and frequencies can then be computed from the resulting generalized mass and stiffness matrices for that mass configuration. The rigid body mass properties of the vehicle are derived from the integrated vehicle model. The coupling terms between the vehicle rigid body motion and elastic deformation are also updated from the constant system shape functions and the integrated vehicle model. This approach was first used to analyze variable mass spinning beams and then prototyped into a generic dynamics simulation engine. The resulting code was tested against Crew Launch Vehicle (CLV-)class problems worked in the TREETOPS simulation package and by Wilson [2]. The Ares I System Integration Laboratory (SIL) is currently being developed at the Marshall Space Flight Center (MSFC) to test vehicle avionics hardware and software in a hardware-in-the-loop (HWIL) environment and certify that the integrated system is prepared for flight. The Ares I SIL utilizes the Ares Real-Time Environment for Modeling, Integration, and Simulation (ARTEMIS) tool to simulate the launch vehicle and stimulate avionics hardware. Due to the presence of vehicle control system filters and the thrust oscillation suppression system, which are tuned to the structural characteristics of the vehicle, ARTEMIS must incorporate accurate structural models of the Ares I launch vehicle. The ARTEMIS core dynamics simulation models the highly coupled nature of the vehicle flexible body dynamics, propellant slosh, and vehicle nozzle inertia effects combined with mass and flexible body properties that vary significant with time during the flight. All forces that act on the vehicle during flight must be simulated, including deflected engine thrust force, spatially distributed aerodynamic forces, gravity, and reaction control jet thrust forces. These forces are used to excite an integrated flexible vehicle, slosh, and nozzle dynamics model for the vehicle stack that simulates large rigid body translations and rotations along with small elastic deformations. Highly effective matrix math operations on a distributed, threaded high-performance simulation node allow ARTEMIS to retain up to 30 modes of flex for real-time simulation. Stage elements that separate from the stack during flight are propagated as independent rigid six degrees of freedom (6DOF) bodies. This paper will present the formulation of the resulting equations of motion, solutions to example problems, and describe the resulting dynamics simulation engine within ARTEMIS.

Tobbe, Patrick A.↗

The persistent shadow of the supermassive black hole of M87. II. Model comparisons and theoretical interpretations

The Event Horizon Telescope (EHT) observation of M87∗ in 2018 has revealed a ring with a diameter that is consistent with the 2017 observation. The brightest part of the ring is shifted to the southwest from the southeast. In this paper, we provide theoretical interpretations for the multi-epoch EHT observations for M87∗ by comparing a new general relativistic magnetohydrodynamics model image library with the EHT observations for M87∗ in both 2017 and 2018. The model images include aligned and tilted accretion with parameterized thermal and nonthermal synchrotron emission properties. The 2018 observation again shows that the spin vector of the M87∗ supermassive black hole is pointed away from Earth. A shift of the brightest part of the ring during the multi-epoch observations can naturally be explained by the turbulent nature of black hole accretion, which is supported by the fact that the more turbulent retrograde models can explain the multi-epoch observations better than the prograde models. The EHT data are inconsistent with the tilted models in our model image library. Assuming that the black hole spin axis and its large-scale jet direction are roughly aligned, we expect the brightest part of the ring to be most commonly observed 90 deg clockwise from the forward jet. This prediction can be statistically tested through future observations.

79 ASTRONOMY AND ASTROPHYSICS↗