Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “memory usage”

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 145 records · Page 8

End-to-end GPU acceleration of low-order-refined preconditioning for high-order finite element discretizations

In this article, we present algorithms and implementations for the end-to-end GPU acceleration of matrix-free low-order-refined preconditioning of high-order finite element problems. The methods described here allow for the construction of effective preconditioners for high-order problems with optimal memory usage and computational complexity. The preconditioners are based on the construction of a spectrally equivalent low-order discretization on a refined mesh, which is then amenable to, for example, algebraic multigrid preconditioning. The constants of equivalence are independent of mesh size and polynomial degree. For vector finite element problems in H(curl) and H(div) (e.g., for electromagnetic or radiation diffusion problems), a specially constructed interpolation–histopolation basis is used to ensure fast convergence. Detailed performance studies are carried out to analyze the efficiency of the GPU algorithms. The kernel throughput of each of the main algorithmic components is measured, and the strong and weak parallel scalability of the methods is demonstrated. The different relative weighting and significance of the algorithmic components on GPUs and CPUs is discussed. Results on problems involving adaptively refined nonconforming meshes are shown, and the use of the preconditioners on a large-scale magnetic diffusion problem using all spaces of the finite element de Rham complex is illustrated.

97 MATHEMATICS AND COMPUTING↗

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↗

Pseudonymized User-Perspective Summit Login Node Data for 2020 and 2021

This dataset contains hourly snapshot data from each of the 5 login nodes of the Summit supercomputer at Oak Ridge Leadership Computing Facility (OLCF) over a period of 2 years, starting January 2020 and ending after December 2021. The snapshots include lists of currently logged-in users, CPU and memory usage, status of users' batch jobs, and disk usage statistics. Usernames, project identifiers, and file paths have been pseudonymized in order to allow studies of user behavior without divulging Personally Identifiable Information (PII).

97 MATHEMATICS AND COMPUTING↗

Demonstration of MOOSE-based Griffin reactor physics, code for heterogeneous lead-cooled fast reactor analysis

The MOOSE-based reactor physics code Griffin was assessed on a heterogeneous pin-resolved model of a prototype lead-cooled fast reactor assembly. This model was developed in preparation for future use in MOOSE-based multiphysics calculations for computing hot channel factors. Heterogeneous multigroup cross sections were prepared using the fast reactor multi-group cross section processing code MC{sup 2}-3 using a two-step method. Griffin simulations were performed using the DFEM-SN solver on 576 cores on Argonne's LCRC cluster. Diffusion-based acceleration methods were applied (NDA and CMFD). Reference solutions were generated with continuous energy MCNP and the hybrid MOC/finite element solver PROTEUS-MOC for code-to-code comparison. Space-angle convergence studies were conducted to observe convergence in k-eigenvalue and axial pin power distributions. The fully resolved Griffin calculation was within 68 pcm of the MCNP eigenvalue and exhibited max 1.2% relative error in the axial pin power distribution. Griffin produced nearly identical results to PROTEUS-MOC when using the same 9-group multigroup cross-section set. Griffin demonstrated favorable scaling in wall-clock time and memory usage when using diffusion-based acceleration methods. Griffin is capable of simulating the pin-resolved heterogeneous LFR assembly with good accuracy and performance, and is suitable for future use in coupled high-fidelity hot channel factor simulations. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Nuclear Energy Advanced Modeling and Simulation (NEAMS) waste Integrated Performance and Safety Codes (IPSC) : gap analysis for high fidelity and performance assessment code development.

This report describes a gap analysis performed in the process of developing the Waste Integrated Performance and Safety Codes (IPSC) in support of the U.S. Department of Energy (DOE) Office of Nuclear Energy Advanced Modeling and Simulation (NEAMS) Campaign. The goal of the Waste IPSC is to develop an integrated suite of computational modeling and simulation capabilities to quantitatively assess the long-term performance of waste forms in the engineered and geologic environments of a radioactive waste storage or disposal system. The Waste IPSC will provide this simulation capability (1) for a range of disposal concepts, waste form types, engineered repository designs, and geologic settings, (2) for a range of time scales and distances, (3) with appropriate consideration of the inherent uncertainties, and (4) in accordance with rigorous verification, validation, and software quality requirements. The gap analyses documented in this report were are performed during an initial gap analysis to identify candidate codes and tools to support the development and integration of the Waste IPSC, and during follow-on activities that delved into more detailed assessments of the various codes that were acquired, studied, and tested. The current Waste IPSC strategy is to acquire and integrate the necessary Waste IPSC capabilities wherever feasible, and develop only those capabilities that cannot be acquired or suitably integrated, verified, or validated. The gap analysis indicates that significant capabilities may already exist in the existing THC codes although there is no single code able to fully account for all physical and chemical processes involved in a waste disposal system. Large gaps exist in modeling chemical processes and their couplings with other processes. The coupling of chemical processes with flow transport and mechanical deformation remains challenging. The data for extreme environments (e.g., for elevated temperature and high ionic strength media) that are needed for repository modeling are severely lacking. In addition, most of existing reactive transport codes were developed for non-radioactive contaminants, and they need to be adapted to account for radionuclide decay and in-growth. The accessibility to the source codes is generally limited. Because the problems of interest for the Waste IPSC are likely to result in relatively large computational models, a compact memory-usage footprint and a fast/robust solution procedure will be needed. A robust massively parallel processing (MPP) capability will also be required to provide reasonable turnaround times on the analyses that will be performed with the code. A performance assessment (PA) calculation for a waste disposal system generally requires a large number (hundreds to thousands) of model simulations to quantify the effect of model parameter uncertainties on the predicted repository performance. A set of codes for a PA calculation must be sufficiently robust and fast in terms of code execution. A PA system as a whole must be able to provide multiple alternative models for a specific set of physical/chemical processes, so that the users can choose various levels of modeling complexity based on their modeling needs. This requires PA codes, preferably, to be highly modularized. Most of the existing codes have difficulties meeting these requirements. Based on the gap analysis results, we have made the following recommendations for the code selection and code development for the NEAMS waste IPSC: (1) build fully coupled high-fidelity THCMBR codes using the existing SIERRA codes (e.g., ARIA and ADAGIO) and platform, (2) use DAKOTA to build an enhanced performance assessment system (EPAS), and build a modular code architecture and key code modules for performance assessments. The key chemical calculation modules will be built by expanding the existing CANTERA capabilities as well as by extracting useful components from other existing codes.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

INDDGO: Integrated Network Decomposition & Dynamic programming for Graph Optimization

It is well-known that dynamic programming algorithms can utilize tree decompositions to provide a way to solve some NP -hard problems on graphs where the complexity is polynomial in the number of nodes and edges in the graph, but exponential in the width of the underlying tree decomposition. However, there has been relatively little computational work done to determine the practical utility of such dynamic programming algorithms. We have developed software to construct tree decompositions using various heuristics and have created a fast, memory-efficient dynamic programming implementation for solving maximum weighted independent set. We describe our software and the algorithms we have implemented, focusing on memory saving techniques for the dynamic programming. We compare the running time and memory usage of our implementation with other techniques for solving maximum weighted independent set, including a commercial integer programming solver and a semi-definite programming solver. Our results indicate that it is possible to solve some instances where the underlying decomposition has width much larger than suggested by the literature. For certain types of problems, our dynamic programming code runs several times faster than these other methods.

Groër, Christopher S.↗

Review of Graph Databases for Big Data Dynamic Entity Scoring

Modeling data as a graph enables users to quickly analyze networked phenomenon, such as social network-based marketing data (e.g., linking entities in social media based on their friends and their "likes", their friend-of-a-friend's "likes", etc.) and scientific data, such as in biology to assess a gene's proximity in a graph to particular disease phenotype. Numerous purpose-built graph databases (DBs) called “NoSQL” or “Not Only SQL” DBs are emerging to support this type of graph analysis. These DBs differ from traditional relational DBs in that they are optimized to improve storage and query performance across large, complex graphs. This document reviews the state of the art in graph DBs and identifies potential DBs that show promise for dynamic entity-scoring algorithms. Dynamic entity scoring is a new area of research that allows analysis of data within graphs using undirected methods to weigh the edges of the graph based on the proximity or ambient influence of near-by nodes. The EntityScore (eScore) methodology developed at Lawrence Livermore National Laboratory (LLNL) is one application of these emerging graph-analysis algorithms to help analysts identify and prioritize entities, based on complex, analyst-defined weighting criteria. Our review examines a wide range of different, well-known graph DBs in use since at least 2010 and identifies unique advantages and disadvantages that may impact eScore-specific requirements. Specifically, we review data structures employed, query features employed, load times, query calculation times, memory usage and several other important attributes reported in the literature. We limited our considerations for eScore to Blueprints-compliant graph DBs (https://github.com/tinkerpop/blueprints/) because they provide a standard, java application programming interface (API) based on the property graph model.

42 ENGINEERING↗

Predictive Engineering Tools for Injection-molded Long-Carbon-Fiber Thermoplastic Composites - FY 2014 Third Quarterly Report

This report describes the technical progresses made during the third quarter of FY 2014: 1) Autodesk introduced the options for fiber inlet condition to the 3D solver. These options are already available in the mid-plane/dual domain solver. 2) Autodesk improved the accuracy of 3D fiber orientation calculation around the gate. 3) Autodesk received consultant services from Prof. C.L. Tucker at the University of Illinois on the implementation of the reduced order model for fiber length, and discussed with Prof. Tucker the methods to reduce memory usage. 4) PlastiComp delivered to PNNL center-gated and edge-fan-gated 20-wt% to 30-wt% LCF/PP and LCF/PA66 (7”x7”x1/8”) plaques molded by the in-line direct injection molding (D-LFT) process. 5) PlastiComp molded ASTM tensile, flexural and impact bars under the same D-LFT processing conditions used for plaques for Certification of Assessment and ascertaining the resultant mechanical properties. 6) Purdue developed a new polishing routine, utilizing the automated polishing machine, to reduce fiber damage during surface preparation. 7) Purdue used a marker-based watershed segmentation routine, in conjunction with a hysteresis thresholding technique, for fiber segmentation during fiber orientation measurement. 8) Purdue validated Purdue’s fiber orientation measurement method using the previous fiber orientation data obtained from the Leeds machine and manually measured data by the University of Illinois. 9) PNNL conducted ASMI mid-plane analyses for a 30wt% LCF/PP plaque and compared the predicted fiber orientations with the measured data provided by Purdue University at the selected locations on this plaque. 10) PNNL put together the DOE 2014 Annual Merit Review (AMR) presentation with the team and presented it at the AMR meetings on June 17, 2014. 11) PNNL built ASMI dual domain models for the Toyota complex part and commenced mold filling analyses of the complex part with different wall thicknesses in order to support part molding. 12) Toyota and Magna discussed with PNNL on tool modification for molding the complex part. Toyota sent the CAD files of the complex part to PNNL to build ASMI models of the part for mold filling analysis to provide guidance to tooling and part molding.

Injection molding, long-fiber thermoplastic, carbo↗

Techniques for Automated Performance Analysis

The performance of a particular HPC code depends on a multitude of variables, including compiler selection, optimization flags, OpenMP pool size, file system load, memory usage, MPI configuration, etc. As a result of this complexity, current predictive models have limited applicability, especially at scale. We present a formulation of scientific codes, nodes, and clusters that reduces complex performance analysis to well-known mathematical techniques. Building accurate predictive models and enhancing our understanding of scientific codes at scale is an important step towards exascale computing.

97 MATHEMATICS AND COMPUTING↗

Parallel Block Structured Adaptive Mesh Refinement on Graphics Processing Units

Block-structured adaptive mesh refinement is a technique that can be used when solving partial differential equations to reduce the number of zones necessary to achieve the required accuracy in areas of interest. These areas (shock fronts, material interfaces, etc.) are recursively covered with finer mesh patches that are grouped into a hierarchy of refinement levels. Despite the potential for large savings in computational requirements and memory usage without a corresponding reduction in accuracy, AMR adds overhead in managing the mesh hierarchy, adding complex communication and data movement requirements to a simulation. In this paper, we describe the design and implementation of a native GPU-based AMR library, including: the classes used to manage data on a mesh patch, the routines used for transferring data between GPUs on different nodes, and the data-parallel operators developed to coarsen and refine mesh data. We validate the performance and accuracy of our implementation using three test problems and two architectures: an eight-node cluster, and over four thousand nodes of Oak Ridge National Laboratory’s Titan supercomputer. Our GPU-based AMR hydrodynamics code performs up to 4.87× faster than the CPU-based implementation, and has been scaled to over four thousand GPUs using a combination of MPI and CUDA.

97 MATHEMATICS AND COMPUTING↗

Limited-memory adaptive snapshot selection for proper orthogonal decomposition

Reduced order models are useful for accelerating simulations in many-query contexts, such as optimization, uncertainty quantification, and sensitivity analysis. However, offline training of reduced order models can have prohibitively expensive memory and floating-point operation costs in high-performance computing applications, where memory per core is limited. To overcome this limitation for proper orthogonal decomposition, we propose a novel adaptive selection method for snapshots in time that limits offline training costs by selecting snapshots according an error control mechanism similar to that found in adaptive time-stepping ordinary differential equation solvers. The error estimator used in this work is related to theory bounding the approximation error in time of proper orthogonal decomposition-based reduced order models, and memory usage is minimized by computing the singular value decomposition using a single-pass incremental algorithm. Results for a viscous Burgers’ test problem demonstrate convergence in the limit as the algorithm error tolerances go to zero; in this limit, the full order model is recovered to within discretization error. The resulting method can be used on supercomputers to generate proper orthogonal decomposition-based reduced order models, or as a subroutine within hyperreduction algorithms that require taking snapshots in time, or within greedy algorithms for sampling parameter space.

97 MATHEMATICS AND COMPUTING↗

Adaptive Mesh Refinement in the Fastlane

Adaptive Mesh Refinement (AMR) allows us to dramatically reduce memory usage by dynamically distributing resources to areas of physical significance.

97 MATHEMATICS AND COMPUTING↗

Fault Resilient Domain Decomposition Preconditioner for PDEs

The move towards extreme-scale computing platforms challenges scientific simulations in many ways. Given the recent tendencies in computer architecture development, one needs to reformulate legacy codes in order to cope with large amounts of communication, system faults and requirements of low-memory usage per core. In this work, we develop a novel framework for solving partial differential equations (PDEs) via domain decomposition that reformulates the solution as a state-of-knowledge with a probabilistic interpretation. Such reformulation allows resiliency with respect to potential faults without having to apply fault detection, avoids unnecessary communication and is generally well-positioned for rigorous uncertainty quantification studies that target improvements of predictive fidelity of scientific models. We demonstrate our algorithm for one-dimensional PDE examples where artificial faults have been implemented as bit-flips in the binary representation of subdomain solutions.

97 MATHEMATICS AND COMPUTING↗

Evaluation of Accuracy and Performance of a Bilinear Interpolator with Dense EOS Tables in xRage

This study investigates the performance and accuracy of simulations using air and deuterium equation-of-state (EOS) tables that have been densified coupled with linear interpolation, as opposed to the default rational interpolation used in EOSPAC. EOS tables that are densified by adding in more temperature and density points that have been calculated by the rational function interpolator are able to improve the accuracy of the linear interpolator by over 50%. There was no change in runtime performance for a simulation using a dense EOS table with linear interpolation. In a memory limited case, dense EOS tables had no apparent effect on the simulation runtime or memory usage.

97 MATHEMATICS AND COMPUTING↗

Thin-Threads: An Approach for History-Based Monte Carlo on GPUs

A graphics processing unit (GPU) has become a core technology for modern supercomputers. Applications that once ran on supercomputers are being forced to make significant changes to their designs to utilize these new machines. This paper introduces the concept of Thin-Threads as a method for history-based Monte Carlo transport applications on GPUs. The key principles behind Thin-Threads are light memory usage and communication and managing data race issues via atomics. We show that we can achieve a 10x speedup when moving from the traditional method to Thin-Threads on GPUs. Additionally, we demonstrate the viability of the Thin-Threads model at scale for GPU and CPU platforms.

97 MATHEMATICS AND COMPUTING↗

Sub-Libraries in ORIGEN

Libraries produced for the ORIGEN depletion code have the ability to label nuclides with a sublibrary integer tag. While this is useful for tracking different sources of the nuclides, it adds some complexity, increasing memory usage and computational effort. This report documents the effort of making the sublibraries optional in an ORIGEN calculation, with a focus on the creation of libraries using the LibraryBuilder application programming interface (API). ORIGEN will gain the ability to either use the normal sublibrary set, a reduced, unique, sublibrary set, or no sublibrary at all.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

VeraShift Usability Updates for Production Releases of VERA for Ex-core Calculations

The recently developed capability to perform ex-core calculations in the Virtual Environment for Reactor Applications (VERA) with Shift is now available in production releases of VERA. To support these production releases, updates to the VeraShift interface are needed continually. The updates completed under this milestone are documented in this report, including improvements to memory usage and code robustness, as well as the resolution of issues with unexpected behavior that arose during testing.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Using Likwid and Byfl to Benchmark Hardware Performance

This paper outlines a benchmarking study conducted during my internship at LANL, focusing on CPU (Computer Processing Unit) and program performance assessment. The primary goal was to gather memory access data using three methods across five polybench kernels The data gathered would then be used to compare and contrast to one another and calculate operational intensity for performance comparisons. Benchmarking tools like Byfl and Likwid were employed, with Byfl offering hardware-independent data through LLVM compiler communication and Likwid directly interacting with computer hardware. The study considered various benchmarking factors, including optimization levels, Big O notation ((n)), CPU diversity and specific kernel equations. Big O notation was utilized to simplify code complexity, with detailed breakdwons of operations and memory components for each polybench application. Specific O(n) equations enabled nuanced kernel compariosns, facilitating the identification of performance variations. CPU efficiency assessments were conducted using Likwid tests on two CPUs. The central focus on code optimization aimed at achieving higher speeds and reduced memory usage through streamlined code. Future work propsoes creating a roofline model, synthesizing benchmarking data into a comprehensive data graph to assist in optimizing code and improving hardware performance. The potential impact on the laboratory or national mission was underscored, emphasizing the importance of optimizing applications and hardware to conserve resources and accelerate program execution. The specific relevance to LANL’s operations in math-intensive fields such as Nuclear Fission, Space Exploration, and Nanotechnology highlights the necessity of efficient benchmarking for resource conservation and proram speed. Overall, this study contributes to the understanding of CPU and program performance, providing insights for future optimization efforts in a laboratory setting

97 MATHEMATICS AND COMPUTING↗