Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Distributed Memory Parallel”

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 685 records · Page 38

MPI, HPF or OpenMP: A Study with the NAS Benchmarks

Porting applications to new high performance parallel and distributed platforms is a challenging task. Writing parallel code by hand is time consuming and costly, but the task can be simplified by high level languages and would even better be automated by parallelizing tools and compilers. The definition of HPF (High Performance Fortran, based on data parallel model) and OpenMP (based on shared memory parallel model) standards has offered great opportunity in this respect. Both provide simple and clear interfaces to language like FORTRAN and simplify many tedious tasks encountered in writing message passing programs. In our study we implemented the parallel versions of the NAS Benchmarks with HPF and OpenMP directives. Comparison of their performance with the MPI implementation and pros and cons of different approaches will be discussed along with experience of using computer-aided tools to help parallelize these benchmarks. Based on the study,potentials of applying some of the techniques to realistic aerospace applications will be presented

Jin, Hao-Qiang↗

CAST: Contraction Algorithm for Symmetric Tensors

Tensor contractions represent the most compute-intensive core kernels in ab initio computational quantum chemistry and nuclear physics. Symmetries in these tensor contractions makes them difficult to load balance and scale to large distributed systems. In this paper, we develop an efficient and scalable algorithm to contract symmetric tensors. We introduce a novel approach that avoids data redistribution in contracting symmetric tensors while also avoiding redundant storage and maintaining load balance. We present experimental results on two parallel supercomputers for several symmetric contractions that appear in the CCSD quantum chemistry method. We also present a novel approach to tensor redistribution that can take advantage of parallel hyperplanes when the initial distribution has replicated dimensions, and use collective broadcast when the final distribution has replicated dimensions, making the algorithm very efficient.

distributed memory↗

CID-720 aircraft high-environment flight instrumentation system

The high-environment flight instrumentation system was designed to acquire Langley's structural response data during the full scale transport-controlled impact demonstration test. There was only one opportunity for data acquisition. Thus, a high reliability and crashworthy design approach was implemented. The approach featured multi-level redundancy and a vigorous quality assurance testing program. Complying with an accelerated schedule, the instrumentation system was developed, tested and shipped within 18 months to Dryden Flight Research Facility. The flight instrumentation system consists of two autonomous data systems, DAS #1 and #2, and an excellent checkout subsystem. Each data system is partitioned into four pallets. The system was designed to operate on manned and unmanned flights. There are 176 data channels per data system. These channels are sequentially sampled and encoded into 1 megabit/sec pulse code modulation (PCM) data signal. To increase the probability of success, a special PCM distribution subsystem was developed. This subsystem distributes the PCM signal to two transmitters, one delay memory, and eight recorder tracks. The data on four of these trackes was digitally delayed approximately 300 msec to maximize data acquisition during impact. Therefore each data system's data is redundantly recorded onboard and on the ground. There are two time code generators. Parallel time from each is encoded into both data systems. Serial time from each is redundantly recorded on both onboard recorders. Instrumentation power is independent of aircraft power and self-contained.

Calloway, R. S.↗

HPC Node Performance and Energy Modeling with the Co-Location of Applications

Multicore processors have become an integral part of modern large-scale and high-performance parallel and distributed computing systems. This advance became necessary as the demand for increased system performance has exceeded the limits that single core processors can provide. Unfortunately, applications co-located on multicore processors can suffer from decreased performance and increased dynamic energy use as a result of access to shared resources, such as memory. Consequently, it is increasingly important to characterize the performance of applications that execute on these architectures. This work investigates some of the disadvantages of co-location, and presents a methodology for building models capable of utilizing varying amounts of information about a target application and its co-located applications to make predictions about the target application’s execution time and the system’s energy use under arbitrary co-locations of a wide range of application types. The proposed methodology is validated on three different server class Intel Xeon multicore processors using eleven applications from two scientific benchmark suites. The model’s utility for scheduling is also demonstrated in a simulated large-scale high-performance computing environment through the creation of a co-location aware scheduling heuristic. This heuristic demonstrates that scheduling using information generated with the proposed modeling methodology is capable of making significant improvements over a scheduling heuristic that is naive to co-location interference.

performance modeling, energy modeling, resource ma↗

Enabling In-situ Execution of Coupled Scientific Workflow on Multi-core Platform

Emerging scientific application workflows are composed of heterogeneous coupled component applications that simulate different aspects of the physical phenomena being modeled, and that interact and exchange significant volumes of data at runtime. With the increasing performance gap between on-chip data sharing and off-chip data transfers in current systems based on multicore processors, moving large volumes of data using communication network fabric can significantly impact performance. As a result, minimizing the amount of inter-application data exchanges that are across compute nodes and use the network is critical to achieving overall application performance and system efficiency. In this paper, we investigate the in-situ execution of the coupled components of a scientific application workflow so as to maximize on-chip exchange of data. Specifically, we present a distributed data sharing and task execution framework that (1) employs data-centric task placement to map computations from the coupled applications onto processor cores so that a large portion of the data exchanges can be performed using the intra-node shared memory, (2) provides a shared space programming abstraction that supplements existing parallel programming models (e.g., message passing) with specialized one-sided asynchronous data access operators and can be used to express coordination and data exchanges between the coupled components. We also present the implementation of the framework and its experimental evaluation on the Jaguar Cray XT5 at Oak Ridge National Laboratory.

Computer Science↗

High order parallel numerical schemes for solving incompressible flows

The use of parallel computers for numerically solving flow fields has gained much importance in recent years. This paper introduces a new high order numerical scheme for computational fluid dynamics (CFD) specifically designed for parallel computational environments. A distributed MIMD system gives the flexibility of treating different elements of the governing equations with totally different numerical schemes in different regions of the flow field. The parallel decomposition of the governing operator to be solved is the primary parallel split. The primary parallel split was studied using a hypercube like architecture having clusters of shared memory processors at each node. The approach is demonstrated using examples of simple steady state incompressible flows. Future studies should investigate the secondary split because, depending on the numerical scheme that each of the processors applies and the nature of the flow in the specific subdomain, it may be possible for a processor to seek better, or higher order, schemes for its particular subcase.

Lin, Avi↗

Hierarchical resilience with lightweight threads.

This paper proposes methodology for providing robustness and resilience for a highly threaded distributed- and shared-memory environment based on well-defined inputs and outputs to lightweight tasks. These inputs and outputs form a failure 'barrier', allowing tasks to be restarted or duplicated as necessary. These barriers must be expanded based on task behavior, such as communication between tasks, but do not prohibit any given behavior. One of the trends in high-performance computing codes seems to be a trend toward self-contained functions that mimic functional programming. Software designers are trending toward a model of software design where their core functions are specified in side-effect free or low-side-effect ways, wherein the inputs and outputs of the functions are well-defined. This provides the ability to copy the inputs to wherever they need to be - whether that's the other side of the PCI bus or the other side of the network - do work on that input using local memory, and then copy the outputs back (as needed). This design pattern is popular among new distributed threading environment designs. Such designs include the Barcelona STARS system, distributed OpenMP systems, the Habanero-C and Habanero-Java systems from Vivek Sarkar at Rice University, the HPX/ParalleX model from LSU, as well as our own Scalable Parallel Runtime effort (SPR) and the Trilinos stateless kernels. This design pattern is also shared by CUDA and several OpenMP extensions for GPU-type accelerators (e.g. the PGI OpenMP extensions).

43 PARTICLE ACCELERATORS↗

Technology and future ground processing systems

Land-observing satellites with multiple thematic mappers will produce data at rates of 100 to 300 Mbps. When coupled with a high daily scene production rate, these rates will require new approaches to ground processing. Consideration is given here to future downlink rates and data volumes, and requirements peculiar to the future user community are discussed. The advanced technologies required to attain an operational system in the years 1985-1990 are considered, together with advances foreseen in communications, mass storage, bulk memories, and data processing. Using advanced devices, a centralized data processing system capable of handling the 100 Mbps data rate is described. New approaches, among them a parallel pipelined calibration front-end, real-time browse image production, a high bandwidth optical disk archive, regional image broadcast and massively parallel product production, are considered. A distributed system capable of handling the 300 Mbps data rate is then described. Designs for a hub system and a regional processing center are presented.

Wood, B. J.↗

SPARC: Demonstrate burst-buffer-based checkpoint/restart on ATS-1.

Recent high-performance computing (HPC) platforms such as the Trinity Advanced Technology System (ATS-1) feature burst buffer resources that can have a dramatic impact on an application’s I/O performance. While these non-volatile memory (NVM) resources provide a new tier in the storage hierarchy, developers must find the right way to incorporate the technology into their applications in order to reap the benefits. Similar to other laboratories, Sandia is actively investigating ways in which these resources can be incorporated into our existing libraries and workflows without burdening our application developers with excessive, platform-specific details. This FY18Q1 milestone summaries our progress in adapting the Sandia Parallel Aerodynamics and Reentry Code (SPARC) in Sandia’s ATDM program to leverage Trinity’s burst buffers for checkpoint/restart operations. We investigated four different approaches with varying tradeoffs in this work: (1) simply updating job script to use stage-in/stage out burst buffer directives, (2) modifying SPARC to use LANL’s hierarchical I/O (HIO) library to store/retrieve checkpoints, (3) updating Sandia’s IOSS library to incorporate the burst buffer in all meshing I/O operations, and (4) modifying SPARC to use our Kelpie distributed memory library to store/retrieve checkpoints. Team members were successful in generating initial implementation for all four approaches, but were unable to obtain performance numbers in time for this report (reasons: initial problem sizes were not large enough to stress I/O, and SPARC refactor will require changes to our code). When we presented our work to the SPARC team, they expressed the most interest in the second and third approaches. The HIO work was favored because it is lightweight, unobtrusive, and should be portable to ATS-2. The IOSS work is seen as a long-term solution, and is favored because all I/O work (including checkpoints) can be deferred to a single library.

97 MATHEMATICS AND COMPUTING↗

Modeling the asynchronous Jacobi method without communication delays

Asynchronous iterative methods for solving linear systems are gaining renewed interest due to the high cost of synchronization points in massively parallel codes. Historically, theory on asynchronous iterative methods has focused on asymptotic behavior, while the transient behavior remains poorly understood. In this paper, we study a model of the asynchronous Jacobi method without communication delays, which we call simplified asynchronous Jacobi. Simplified asynchronous Jacobi can be used to model asynchronous Jacobi implemented in shared memory or distributed memory with fast communication networks. Our analysis uses the idea of a propagation matrix, which is similar in concept to an iteration matrix. We show that simplified asynchronous Jacobi can continue to reduce the residual when some processes are slower than other processes. We also show that simplified asynchronous Jacobi can converge when synchronous Jacobi does not. We verify our analysis of simplified asynchronous Jacobi using results from asynchronous Jacobi implemented in shared and distributed memory.

97 MATHEMATICS AND COMPUTING↗

Parallelization of irregularly coupled regular meshes

Regular meshes are frequently used for modeling physical phenomena on both serial and parallel computers. One advantage of regular meshes is that efficient discretization schemes can be implemented in a straight forward manner. However, geometrically-complex objects, such as aircraft, cannot be easily described using a single regular mesh. Multiple interacting regular meshes are frequently used to describe complex geometries. Each mesh models a subregion of the physical domain. The meshes, or subdomains, can be processed in parallel, with periodic updates carried out to move information between the coupled meshes. In many cases, there are a relatively small number (one to a few dozen) subdomains, so that each subdomain may also be partitioned among several processors. We outline a composite run-time/compile-time approach for supporting these problems efficiently on distributed-memory machines. These methods are described in the context of a multiblock fluid dynamics problem developed at LaRC.

Chase, Craig↗

Massively Parallel Quantum Chemistry: A high-performance research platform for electronic structure

The Massively Parallel Quantum Chemistry (MPQC) program is a 30-year-old project that enables facile development of electronic structure methods for molecules for efficient deployment to massively parallel computing architectures. Here, we describe the historical evolution of MPQC’s design into its latest (fourth) version, the capabilities and modular architecture of today’s MPQC, and how MPQC facilitates rapid composition of new methods as well as its state-of-the-art performance on a variety of commodity and high-end distributed-memory computer platforms.

Peng, Chong↗

Supporting 64-bit global indices in Epetra and other Trilinos packages :

The Trilinos Project is an effort to facilitate the design, development, integration and ongoing support of mathematical software libraries within an object-oriented framework. It is intended for large-scale, complex multiphysics engineering and scientific applications [2, 4, 3]. Epetra is one of its basic packages. It provides serial and parallel linear algebra capabilities. Before Trilinos version 11.0, released in 2012, Epetra used the C++ int data-type for storing global and local indices for degrees of freedom (DOFs). Since int is typically 32-bit, this limited the largest problem size to be smaller than approximately two billion DOFs. This was true even if a distributed memory machine could handle larger problems. We have added optional support for C++ long long data-type, which is at least 64-bit wide, for global indices. To save memory, maintain the speed of memory-bound operations, and reduce further changes to the code, the local indices are still 32-bit. We document the changes required to achieve this feature and how the new functionality can be used. We also report on the lessons learned in modifying a mature and popular package from various perspectives design goals, backward compatibility, engineering decisions, C++ language features, effects on existing users and other packages, and build integration.

Jhurani, Chetan↗

Optimizing Distributed Training on Frontier for Large Language Models

Large language models (LLMs) have demonstrated remarkable success as foundational models, benefiting various downstream applications through fine-tuning. Loss scaling studies have demonstrated the superior performance of larger LLMs compared to their smaller counterparts. Nevertheless, training LLMs with billions of parameters poses significant challenges and requires considerable computational resources. For example, training a one trillion parameter GPT-style model on 20 trillion tokens requires a staggering 120 million exaflops. This research explores efficient distributed training strategies to extract this computation from Frontier, the world's first exascale supercomputer. We enable and investigate various model and data parallel training techniques, such as tensor parallelism, pipeline parallelism, and sharded data parallelism, to facilitate training a trillion-parameter model on Frontier. We empirically assess these techniques and their associated parameters to determine their impact on memory footprint, communication latency, and GPU's computational efficiency. We analyze the complex interplay among these techniques and find a strategy to combine them to achieve high throughput through hyperparameter tuning. We have identified efficient strategies for training large LLMs of varying sizes through empirical analysis and hyperparameter tuning. For 22 Billion, 175 Billion, and 1 Trillion parameters, we achieved GPU throughputs of 38.38%, 36.14%, and 31.96%, respectively. For the training of the 175 Billion parameter model and the 1 Trillion parameter model, we achieved 100% weak scaling efficiency on 1024 and 3072 Mi250X GPUs, respectively. We also achieved strong scaling efficiencies of 89% and 87% for these two models. We trained these models only tens of iterations instead of training till completion.

Yin, Junqi↗

Asymptotic Behavior of Memristive Circuits

The interest in memristors has risen due to their possible application both as memory units and as computational devices in combination with CMOS. This is in part due to their nonlinear dynamics, and a strong dependence on the circuit topology. We provide evidence that also purely memristive circuits can be employed for computational purposes. In the present paper we show that a polynomial Lyapunov function in the memory parameters exists for the case of DC controlled memristors. Such a Lyapunov function can be asymptotically approximated with binary variables, and mapped to quadratic combinatorial optimization problems. This also shows a direct parallel between memristive circuits and the Hopfield-Little model. In the case of Erdos-Renyi random circuits, we show numerically that the distribution of the matrix elements of the projectors can be roughly approximated with a Gaussian distribution, and that it scales with the inverse square root of the number of elements. This provides an approximated but direct connection with the physics of disordered system and, in particular, of mean field spin glasses. Using this and the fact that the interaction is controlled by a projector operator on the loop space of the circuit. We estimate the number of stationary points of the approximate Lyapunov function and provide a scaling formula as an upper bound in terms of the circuit topology only.

97 MATHEMATICS AND COMPUTING↗

The SGI/Cray T3E: Experiences and Insights

The NASA Goddard Space Flight Center is home to the fifth most powerful supercomputer in the world, a 1024 processor SGI/Cray T3E-600. The original 512 processor system was placed at Goddard in March, 1997 as part of a cooperative agreement between the High Performance Computing and Communications Program's Earth and Space Sciences Project (ESS) and SGI/Cray Research. The goal of this system is to facilitate achievement of the Project milestones of 10, 50 and 100 GFLOPS sustained performance on selected Earth and space science application codes. The additional 512 processors were purchased in March, 1998 by the NASA Earth Science Enterprise for the NASA Seasonal to Interannual Prediction Project (NSIPP). These two "halves" still operate as a single system, and must satisfy the unique requirements of both aforementioned groups, as well as guest researchers from the Earth, space, microgravity, manned space flight and aeronautics communities. Few large scalable parallel systems are configured for capability computing, so models are hard to find. This unique environment has created a challenging system administration task, and has yielded some insights into the supercomputing needs of the various NASA Enterprises, as well as insights into the strengths and weaknesses of the T3E architecture and software. The T3E is a distributed memory system in which the processing elements (PE's) are connected by a low latency, high bandwidth bidirectional 3-D torus. Due to the focus on high speed communication between PE's, the T3E requires PE's to be allocated contiguously per job. Further, jobs will only execute on the user specified number of PE's and PE timesharing is possible but impractical. With a highly varied job mix in both size and runtime of jobs, the resulting scenario is PE fragmentation and an inability to achieve near 100% utilization. SGI/Cray has provided several scheduling and configuration tools to minimize the impact of fragmentation. These tools include PScheD (the political scheduler), GRM (the global resource manager) and NQE (the Network Queuing Environment). Features and impact of these tools will be discussed, as will resulting performance and utilization data. As a distributed memory system, the T3E is designed to be programmed through explicit message passing. Consequently, certain assumptions related to code design are made by the operating system (UNICOS/mk) and its scheduling tools. With the exception of HPF, which does run on the T3E, however poorly, alternative programming styles have the potential to impact the T3E in unexpected and undesirable ways. Several examples will be presented (preceeded with the disclaimer, "Don't try this at home! Violators will be prosecuted!")

Bernard, Lisa Hamet↗

Tuning collective communication for Partitioned Global Address Space programming models

Partitioned Global Address Space (PGAS) languages offer programmers the convenience of a shared memory programming style combined with locality control necessary to run on large-scale distributed memory systems. Even within a PGAS language programmers often need to perform global communication operations such as broadcasts or reductions, which are best performed as collective operations in which a group of threads work together to perform the operation. In this study we consider the problem of implementing collective communication within PGAS languages and explore some of the design trade-offs in both the interface and implementation. In particular, PGAS collectives have semantic issues that are different than in send–receive style message passing programs, and different implementation approaches that take advantage of the one-sided communication style in these languages. We present an implementation framework for PGAS collectives as part of the GASNet communication layer, which supports shared memory, distributed memory and hybrids. The framework supports a broad set of algorithms for each collective, over which the implementation may be automatically tuned. In conclusion, we demonstrate the benefit of optimized GASNet collectives using application benchmarks written in UPC, and demonstrate that the GASNet collectives can deliver scalable performance on a variety of state-of-the-art parallel machines including a Cray XT4, an IBM BlueGene/P, and a Sun Constellation system with InfiniBand interconnect.

97 MATHEMATICS AND COMPUTING↗

Compressed basis GMRES on high-performance graphics processing units

Krylov methods provide a fast and highly parallel numerical tool for the iterative solution of many large-scale sparse linear systems. To a large extent, the performance of practical realizations of these methods is constrained by the communication bandwidth in current computer architectures, motivating the investigation of sophisticated techniques to avoid, reduce, and/or hide the message-passing costs (in distributed platforms) and the memory accesses (in all architectures). This article leverages Ginkgo’s memory accessor in order to integrate a communication-reduction strategy into the (Krylov) GMRES solver that decouples the storage format (i.e., the data representation in memory) of the orthogonal basis from the arithmetic precision that is employed during the operations with that basis. Given that the execution time of the GMRES solver is largely determined by the memory accesses, the cost of the datatype transforms can be mostly hidden, resulting in the acceleration of the iterative step via a decrease in the volume of bits being retrieved from memory. Together with the special properties of the orthonormal basis (whose elements are all bounded by 1), this paves the road toward the aggressive customization of the storage format, which includes some floating-point as well as fixed-point formats with mild impact on the convergence of the iterative process. We develop a high-performance implementation of the “compressed basis GMRES” solver in the Ginkgo sparse linear algebra library using a large set of test problems from the SuiteSparse Matrix Collection. We demonstrate robustness and performance advantages on a modern NVIDIA V100 graphics processing unit (GPU) of up to 50% over the standard GMRES solver that stores all data in IEEE double-precision.

97 MATHEMATICS AND COMPUTING↗