Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Intel”

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 523 records · Page 29

Characterization of Proxy Application Performance on Advanced Architectures. UMT2013, MCB, AMG2013

Three codes were tested at LLNL as part of a Tri-Lab effort to make detailed assessments of several proxy applications on various advanced architectures, with the eventual goal of extending these assessments to codes of programmatic interest running more realistic simulations. Teams from Sandia and Los Alamos tested proxy apps of their own. The focus in this report is on the LLNL codes UMT2013, MCB, and AMG2013. We present weak and strong MPI scaling results and studies of OpenMP efficiency on a large BG/Q system at LLNL, with comparison against similar tests on an Intel Sandy Bridge TLCC2 system. The hardware counters on BG/Q provide detailed information on many aspects of on-node performance, while information from the mpiP tool gives insight into the reasons for the differing scaling behavior on these two different architectures. Results from three more speculative tests are also included: one that exploits NVRAM as extended memory, one that studies performance under a power bound, and one that illustrates the effects of changing the torus network mapping on BG/Q.

97 MATHEMATICS AND COMPUTING↗

Traileka Glacier X-Stack. Final Scientific/Technical Report

The XStack Traleika Glacier (XSTG) project was a three-year research award for exploring a revolutionary exascale-class machine software framework. The XSTG program, including Intel, UC San Diego, Pacific Northwest National Lab, UIUC, Rice University, Reservoir Labs, ET International, and U. Delaware, had major accomplishments, insights, and products resulting from this three-year effort.

42 ENGINEERING↗

Quantum Materials at the Nanoscale - Final Report

The central aim of the Quantum Materials at the Nanoscale (QMN) cluster was to understand and control collective behavior involving the interplay of spins, orbitals, and charges, which governs many scientifically interesting and technologically important phenomena in numerous complex materials. Because these phenomena involve various competing interactions, and influence properties on many different length and energy scales in complex materials, tackling this important area of study motivated a collaborative effort that combined the diverse capabilities of QMN cluster experimentalists, the essential theoretical analysis provided by QMN cluster theorists, and the outstanding facilities and staff of the FSMRL. During the funding period 2007-2014, the DOE cluster grant for the Quantum Materials at the Nanoscale (QMN) cluster supported, at various times, 15 different faculty members (14 in Physics and 1 in Materials Science and Engineering), 7 postdoctoral research associates, and 57 physics and materials science PhD students. 41 of these PhD students have since graduated and have gone on to a variety of advanced technical positions at universities, industries, and national labs: 25 obtained postdoctoral positions at universities (14), industrial labs (2 at IBM), DOE national facilities (3 at Argonne National Laboratory, 1 at Brookhaven National Lab, 1 at Lawrence Berkeley National Lab, and 1 at Sandia National Lab), and other federal facilities (2 at NIST); 13 took various industrial positions, including positions at Intel (5), Quantum Design (1), Lasque Industries (1), Amazon (1), Bloomberg (1), and J.P. Morgan (1). Thus, the QMN grant provided the essential support for training a large number of technically advanced personnel who have now entered key national facilities, industries, and institutions. Additionally, during the period 2007-2015, the QMN cluster produced 159 publications (see pages 14-23), including 23 papers published in Physical Review Letters; 16 papers in Nature, Nature Physics, Nature Materials, or Nature Communications; 4 papers in Science, and 8 papers in Applied Physics Letters. In this report, we provide some key highlights of the collaborative projects in which the QMN cluster members have been involved since 2007.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Programming Models in HPC

These are the slides for a presentation on programming models in HPC, at the Los Alamos National Laboratory's Parallel Computing Summer School. The following topics are covered: Flynn's Taxonomy of computer architectures; single instruction single data; single instruction multiple data; multiple instruction multiple data; address space organization; definition of Trinity (Intel Xeon-Phi is a MIMD architecture); single program multiple data; multiple program multiple data; ExMatEx workflow overview; definition of a programming model, programming languages, runtime systems; programming model and environments; MPI (Message Passing Interface); OpenMP; Kokkos (Performance Portable Thread-Parallel Programming Model); Kokkos abstractions, patterns, policies, and spaces; RAJA, a systematic approach to node-level portability and tuning; overview of the Legion Programming Model; mapping tasks and data to hardware resources; interoperability: supporting task-level models; Legion S3D execution and performance details; workflow, integration of external resources into the programming model.

97 MATHEMATICS AND COMPUTING↗

Closeout Report ARRA supplement to DE-FG02-08ER41546, 03/15/2010 to 03/14/2011 - Advanced Transfer Map Methods for the Description of Particle Beam Dynamics

The ARRA funds were utilized to acquire a cluster of high performance computers, consisting of one Altus 2804 Server based on a Quad AMD Opteron 6174 12C with 4 2.2 GHz nodes of 12 cores each, resulting in 48 directly usable cores; as well as a Relion 1751 Server using an Intel Xeon X5677 consisting of 4 3.46 GHz cores supporting 8 threads. Both systems run the Unix flavor CentOS, which is designed for use without need of updates, which greatly enhances their reliability. The systems are used to operate our COSY INFINITY environment which supports MPI parallelization. The units arrived at MSU in September 2010, and were taken into operation shortly thereafter.

43 PARTICLE ACCELERATORS↗

Summary of Documentation for DYNA3D-ParaDyn's Software Quality Assurance Regression Test Problems

The Software Quality Assurance (SQA) regression test suite for DYNA3D (Zywicz and Lin, 2015) and ParaDyn (DeGroot, et al., 2015) currently contains approximately 600 problems divided into 21 suites, and is a required component of ParaDyn’s SQA plan (Ferencz and Oliver, 2013). The regression suite allows developers to ensure that software modifications do not unintentionally alter the code response. The entire regression suite is run prior to permanently incorporating any software modification or addition. When code modifications alter test problem results, the specific cause must be determined and fully understood before the software changes and revised test answers can be incorporated. The regression suite is executed on LLNL platforms using a Python script and an associated data file. The user specifies the DYNA3D or ParaDyn executable, number of processors to use, test problems to run, and other options to the script. The data file details how each problem and its answer extraction scripts are executed. For each problem in the regression suite there exists an input deck, an eight-processor partition file, an answer file, and various extraction scripts. These scripts assemble a temporary answer file in a specific format from the simulation results. The temporary and stored answer files are compared to a specific level of numerical precision, and when differences are detected the test problem is flagged as failed. Presently, numerical results are stored and compared to 16 digits. At this accuracy level different processor types, compilers, number of partitions, etc. impact the results to various degrees. Thus, for consistency purposes the regression suite is run with ParaDyn using 8 processors on machines with a specific processor type (currently the Intel Xeon E5530 processor). For non-parallel regression problems, i.e., the two XFEM problems, DYNA3D is used instead. When environments or platforms change, executables using the current source code and the new resource are created and the regression suite is run. If differences in answers arise, the new answers are retained provided that the differences are inconsequential. This bootstrap approach allows the test suite answers to evolve in a controlled manner with a high level of confidence. Developers also run the entire regression suite with (serial) DYNA3D. While these results normally differ from the stored (parallel) answers, abnormal termination or wildly different values are strong indicators of potential issues.

42 ENGINEERING↗

HPC Institutional Computing Project: W15_lesreactiveflow KIVA-hpFE Development: A Robust and Accurate Engine Modeling Software

KIVA-hpFE is a high performance computer software for solving the physics of multi-species and multiphase turbulent reactive flow in complex geometries having immersed moving parts. The code is written in Fortran 90/95 and can be used on any computer platform with any popular complier. The code is in two versions, a serial version and a parallel version utilizing MPICH2 type Message Passing Interface (MPI or Intel MPI) for solving distributed domains. The parallel version is at least 30x faster than the serial version and much faster than our previous generation of parallel engine modeling software, by many factors. The 5th generation algorithm construction is a Galerkin type Finite Element Method (FEM) solving conservative momentum, species, and energy transport equations along with two-equation turbulent model k-ω Reynolds Averaged Navier-Stokes (RANS) model and a Vreman type dynamic Large Eddy Simulation (LES) method. The LES method is capable modeling transitional flow from laminar to fully turbulent; therefore, this LES method does not require special hybrid or blending to walls. The FEM projection method also uses a Petrov-Galerkin (P-G) stabilization along with pressure stabilization. We employ hierarchical basis sets, constructed on the fly with enrichment in areas associated with relatively larger error as determined by error estimation methods. In addition, when not using the hp-adaptive module, the code employs Lagrangian basis or shape functions. The shape functions are constructed for hexahedral, prismatic and tetrahedral elements. The software is designed to solve many types of reactive flow problems, from burners to internal combustion engines and turbines. In addition, the formulation allows for direct integration of solid bodies (conjugate heat transfer), as in heat transfer through housings, parts, cylinders. It can also easily be extended to stress modeling of solids, used in fluid structure interactions problems, solidification, porous media modeling and magneto hydrodynamics.

97 MATHEMATICS AND COMPUTING↗

(U) Status of Trinity and Crossroads Systems

(U) This paper provides a general overview of current and future plans for the Advanced Simulation and Computing (ASC) Advanced Technology (AT) systems fielded by the New Mexico Alliance for Computing at Extreme Scale (ACES), a collaboration between Los Alamos Laboratory and Sandia National Laboratories. Additionally, this paper touches on research of technology beyond traditional CMOS. The status of Trinity, ASCs first AT system, and Crossroads, anticipated to succeed Trinity as the third AT system in 2020 will be presented, along with initial performance studies of the Intel Knights Landing Xeon Phi processors, introduced on Trinity. The challenges and opportunities for our production simulation codes on AT systems will also be discussed. Trinity and Crossroads are a joint procurement by ACES and Lawrence Berkeley Laboratory as part of the Alliance for application Performance at EXtreme scale (APEX) http://apex.lanl.gov.

97 MATHEMATICS AND COMPUTING↗

The Portals 4.1 Network Programming Interface

This report presents a specification for the Portals 4 networ k programming interface. Portals 4 is intended to allow scalable, high-performance network communication betwee n nodes of a parallel computing system. Portals 4 is well suited to massively parallel processing and embedded syste ms. Portals 4 represents an adaption of the data movement layer developed for massively parallel processing platfor ms, such as the 4500-node Intel TeraFLOPS machine. Sandia's Cplant cluster project motivated the development of Version 3.0, which was later extended to Version 3.3 as part of the Cray Red Storm machine and XT line. Version 4 is tar geted to the next generation of machines employing advanced network interface architectures that support enh anced offload capabilities.

97 MATHEMATICS AND COMPUTING↗

Models for Total-Dose Radiation Effects in Non-Volatile Memory

The objective of this work is to develop models to predict radiation effects in non- volatile memory: flash memory and ferroelectric RAM. In flash memory experiments have found that the internal high-voltage generators (charge pumps) are the most sensitive to radiation damage. Models are presented for radiation effects in charge pumps that demonstrate the experimental results. Floating gate models are developed for the memory cell in two types of flash memory devices by Intel and Samsung. These models utilize Fowler-Nordheim tunneling and hot electron injection to charge and erase the floating gate. Erase times are calculated from the models and compared with experimental results for different radiation doses. FRAM is less sensitive to radiation than flash memory, but measurements show that above 100 Krad FRAM suffers from a large increase in leakage current. A model for this effect is developed which compares closely with the measurements.

97 MATHEMATICS AND COMPUTING↗

A Note on Compiling Fortran

Fortran modules tend to serialize compilation of large Fortran projects, by introducing dependencies among the source files. If file A depends on file B, (A uses a module defined by B), you must finish compiling B before you can begin compiling A. Some Fortran compilers (Intel ifort, GNU gfortran and IBM xlf, at least) offer an option to ‘‘verify syntax’’, with the side effect of also producing any associated Fortran module files. As it happens, this option usually runs much faster than the object code generation and optimization phases. For some projects on some machines, it can be advantageous to compile in two passes: The first pass generates the module files, quickly; the second pass produces the object files, in parallel. We achieve a 3.8× speedup in the case study below.

97 MATHEMATICS AND COMPUTING↗

Kokkos' Task DAG Capabilities.

This report documents the ASC/ATDM Kokkos deliverable "Production Portable Dy- namic Task DAG Capability." This capability enables applications to create and execute a dynamic task DAG ; a collection of heterogeneous computational tasks with a directed acyclic graph (DAG) of "execute after" dependencies where tasks and their dependencies are dynamically created and destroyed as tasks execute. The Kokkos task scheduler executes the dynamic task DAG on the target execution resource; e.g. a multicore CPU, a manycore CPU such as Intel's Knights Landing (KNL), or an NVIDIA GPU. Several major technical challenges had to be addressed during development of Kokkos' Task DAG capability: (1) portability to a GPU with it's simplified hardware and micro- runtime, (2) thread-scalable memory allocation and deallocation from a bounded pool of memory, (3) thread-scalable scheduler for dynamic task DAG, (4) usability by applications.

97 MATHEMATICS AND COMPUTING↗

Deploy Nalu/Kokkos algorithmic infrastructure with performance benchmarking.

The former Nalu interior heterogeneous algorithm design, which was originally designed to manage matrix assembly operations over all elemental topology types, has been modified to operate over homogeneous collections of mesh entities. This newly templated kernel design allows for removal of workset variable resize operations that were formerly required at each loop over a Sierra ToolKit (STK) bucket (nominally, 512 entities in size). Extensive usage of the Standard Template Library (STL) std::vector has been removed in favor of intrinsic Kokkos memory views. In this milestone effort, the transition to Kokkos as the underlying infrastructure to support performance and portability on many-core architectures has been deployed for key matrix algorithmic kernels. A unit-test driven design effort has developed a homogeneous entity algorithm that employs a team-based thread parallelism construct. The STK Single Instruction Multiple Data (SIMD) infrastructure is used to interleave data for improved vectorization. The collective algorithm design, which allows for concurrent threading and SIMD management, has been deployed for the core low-Mach element- based algorithm. Several tests to ascertain SIMD performance on Intel KNL and Haswell architectures have been carried out. The performance test matrix includes evaluation of both low- and higher-order methods. The higher-order low-Mach methodology builds on polynomial promotion of the core low-order control volume nite element method (CVFEM). Performance testing of the Kokkos-view/SIMD design indicates low-order matrix assembly kernel speed-up ranging between two and four times depending on mesh loading and node count. Better speedups are observed for higher-order meshes (currently only P=2 has been tested) especially on KNL. The increased workload per element on higher-order meshes bene ts from the wide SIMD width on KNL machines. Combining multiple threads with SIMD on KNL achieves a 4.6x speedup over the baseline, with assembly timings faster than that observed on Haswell architecture. The computational workload of higher-order meshes, therefore, seems ideally suited for the many-core architecture and justi es further exploration of higher-order on NGP platforms. A Trilinos/Tpetra-based multi-threaded GMRES preconditioned by symmetric Gauss Seidel (SGS) represents the core solver infrastructure for the low-Mach advection/diffusion implicit solves. The threaded solver stack has been tested on small problems on NREL's Peregrine system using the newly developed and deployed Kokkos-view/SIMD kernels. fforts are underway to deploy the Tpetra-based solver stack on NERSC Cori system to benchmark its performance at scale on KNL machines.

97 MATHEMATICS AND COMPUTING↗

Sparse Matrix-Matrix Multiplication on Multilevel Memory Architectures: Algorithms and Experiments

Architectures with multiple classes of memory media are becoming a common part of mainstream supercomputer deployments. So called multi-level memories offer differing characteristics for each memory component including variation in bandwidth, latency and capacity. This paper investigates the performance of sparse matrix multiplication kernels on two leading highperformance computing architectures — Intel's Knights Landing processor and NVIDIA's Pascal GPU. We describe a data placement method and a chunking-based algorithm for our kernels that exploits the existence of the multiple memory spaces in each hardware platform. We evaluate the performance of these methods w.r.t. standard algorithms using the auto-caching mechanisms Our results show that standard algorithms that exploit cache reuse performed as well as multi-memory-aware algorithms for architectures such as Ki\iLs where the memory subsystems have similar latencies. However, for architectures such as GPUS where memory subsystems differ significantly in both bandwidth and latency, multi-memory-aware methods are crucial for good performance. In addition, our new approaches permit the user to run problems that require larger capacities than the fastest memory of each compute node without depending on the software-managed cache mechanisms.

97 MATHEMATICS AND COMPUTING↗

MFIX-DEM Phi: Performance and Capability Improvements Towards Industrial Grade Open-source DEM Framework with Integrated Uncertainty Quantification

Two major challenges hinder the effective use and adoption of multiphase computational fluid dynamics tools by the industry. The first is the need for significant computational resources, which is inversely proportional to the accuracy of solutions due to computational intensity of the algorithms. The second barrier is assessing the prediction credibility and confidence in the simulation results. In this project, a multi-tiered approach has been proposed under four broad activities to overcome these challenges while addressing all of the objectives outlined in FOA-0001238 through Phases 1 and 2 of the project. The present report consists of the results for only Phase 1, which was the funded performance period. From the start the project, all of the objectives outlined in FOA were addressed through four major activity tasks in an integrated and balanced fashion to improve adoption of MFIX suite of solvers for industrial use. The first task aimed to improve the performance of MFIX-DEM specifically targeting to acquire the peak performance on Intel Xeon and Xeon Phi based systems, which are expected to be one of the primary high-performance computing platforms both affordable and available for the industrial users in the next two to five years. However, due to a number of changes in course of the project, the scope of the performance improvements related task was significantly reduced to avoid duplicate work. Hence, more emphasis was placed on the other three tasks as discussed below. The second task aimed at physical modeling enhancements through implementation of polydispersity capability and validation of heat transfer models in MFIX. An extended verification and validation (V&V) study was performed for the new polydispersity feature implemented in MFIX-DEM both for granular and coupled gas-solid flows. The features of the polydispersity capability and results for an industrially relevant problem were disseminated through journal papers (one published and one under review at the time of writing of the final technical report). As part of the validation efforts, another industrially relevant problem of interest based on rotary drums was studied for several modes of heat transfer and results were presented in conferences. Third task was aimed towards an important and unique contribution of the project, which was to develop a unified uncertainty quantification framework by integrating MFIX-DEM with a graphical user interface (GUI) driven uncertainty quantification (UQ) engine, i.e., MFIX-GUI and PSUADE. The goal was to enable a user with only modest knowledge of statistics to effectively utilize the UQ framework offered with MFIX-DEM Phi to perform UQ analysis routinely. For Phase 1, a proof-of-concept demonstration of the proposed framework was completed and shared. Direct industry involvement was one of the key virtues of this project, which was performed through forth task. For this purpose, even at the proposal stage, the project team received strong interest in the proposed capabilities from two major corporations, which were further expanded throughout Phase 1 and a new collaboration with another major corporation from chemical industry was also initiated. The level of interest received and continued collaboration for the project during Phase 1 clearly shows the relevance and potential impact of the project for the industrial users.

42 ENGINEERING↗

Monthly report for ECP ATDM ST projects (July 2018)

The team worked on supporting compile and build on ATS2 initial deployment systems for SPARC. Conducted performance runs of EMPIRE on Trinity and an initial compile and run of SPARC on Intel Skylake processors.

97 MATHEMATICS AND COMPUTING↗

Large-Scale Simulation of Brain Tissue, Blue Brain Project, EPFL

The initial version of this project described in the awarded proposal aimed at largely focusing on the evaluation of Intel KNL hardware and the development of new scientific use cases such as then eurorobotics and whole brain modeling. However, during the course of the project EPFL has been awarded a high impact INCITE project focused on the development of models to support brain plasticity which is considered as the key toward understanding memory. Considering this very prestigious opportunity a large part of our developments has been refocused on the development of methods to support the plasticity use case which is described in Annex 1 along with the awarded INCITE proposal.

60 APPLIED LIFE SCIENCES↗

The Portals 4.2 Network Programming Interface

This report presents a specification for the Portals 4 network programming interface. Portals 4 is intended to allow scalable, high-performance network communication between nodes of a parallel computing system. Portals 4 is well suited to massively parallel processing and embedded systems. Portals 4 represents an adaption of the data movement layer developed for massively parallel processing platforms, such as the 4500-node Intel TeraFLOPS machine. Sandia's Cplant cluster project motivated the development of Version 3.0, which was later extended to Version 3.3 as part of the Cray Red Storm machine and XT line. Version 4 is targeted to the next generation of machines employing advanced network interface architectures that support enhanced offload capabilities.

97 MATHEMATICS AND COMPUTING↗