Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel computer architecture”

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 379 records · Page 21

Integration experiments and performance studies of a COTS parallel archive system

Current and future Archive Storage Systems have been asked to (a) scale to very high bandwidths, (b) scale in metadata performance, (c) support policy-based hierarchical storage management capability, (d) scale in supporting changing needs of very large data sets, (e) support standard interface, and (f) utilize commercial-off-the-shelf (COTS) hardware. Parallel file systems have been asked to do the same thing but at one or more orders of magnitude faster in performance. Archive systems continue to move closer to file systems in their design due to the need for speed and bandwidth, especially metadata searching speeds such as more caching and less robust semantics. Currently the number of extreme highly scalable parallel archive solutions is very small especially those that will move a single large striped parallel disk file onto many tapes in parallel. We believe that a hybrid storage approach of using COTS components and innovative software technology can bring new capabilities into a production environment for the HPC community much faster than the approach of creating and maintaining a complete end-to-end unique parallel archive software solution. In this paper, we relay our experience of integrating a global parallel file system and a standard backup/archive product with a very small amount of additional code to provide a scalable, parallel archive. Our solution has a high degree of overlap with current parallel archive products including (a) doing parallel movement to/from tape for a single large parallel file, (b) hierarchical storage management, (c) ILM features, (d) high volume (non-single parallel file) archives for backup/archive/content management, and (e) leveraging all free file movement tools in Linux such as copy, move, Is, tar, etc. We have successfully applied our working COTS Parallel Archive System to the current world's first petafiop/s computing system, LANL's Roadrunner machine, and demonstrated its capability to address requirements of future archival storage systems.

99 GENERAL AND MISCELLANEOUS↗

Reproduced Computational Results Report for “Ginkgo: A Modern Linear Operator Algebra Framework for High Performance Computing”

The article titled “Ginkgo: A Modern Linear Operator Algebra Framework for High Performance Computing” by Anzt et al. presents a modern, linear operator centric, C++ library for sparse linear algebra. Experimental results in the article demonstrate that Ginkgo is a flexible and user-friendly framework capable of achieving high-performance on state-of-the-art GPU architectures. In this report, the Ginkgo library is installed and a subset of the experimental results are reproduced. Specifically, the experiment that shows the achieved memory bandwidth of the Ginkgo Krylov linear solvers on NVIDIA A100 and AMD MI100 GPUs is redone and the results are compared to what presented in the published article. Upon completion of the comparison, the published results are deemed reproducible.

97 MATHEMATICS AND COMPUTING↗

Scaling deep learning workloads: NVIDIA DGX-1/Pascal and Intel Knights Landing

Deep Learning (DL) algorithms have become ubiquitous in data analytics. As a result, major computing vendors --- including NVIDIA, Intel, AMD, and IBM --- have architectural road-maps influenced by DL workloads. Furthermore, several vendors have recently advertised new computing products as accelerating large DL workloads. Unfortunately, it is difficult for data scientists to quantify the potential of these different products. This paper provides a performance and power analysis of important DL workloads on two major parallel architectures: NVIDIA DGX-1 (eight Pascal P100 GPUs interconnected with NVLink) and Intel Knights Landing (KNL) CPUs interconnected with Intel Omni-Path or Cray Aries. Our evaluation consists of a cross section of convolutional neural net workloads: CifarNet, AlexNet, GoogLeNet, and ResNet50 topologies using the Cifar10 and ImageNet datasets. The workloads are vendor-optimized for each architecture. Our analysis indicates that although GPUs provide the highest overall performance, the gap can close for some convolutional networks; and the KNL can be competitive in performance/watt. We find that NVLink facilitates scaling efficiency on GPUs. However, its importance is heavily dependent on neural network architecture. Furthermore, for weak-scaling --- sometimes encouraged by restricted GPU memory --- NVLink is less important.

Gawande, Nitin A.↗

XPRESS: eXascale PRogramming Environment and System Software

The XPRESS project is the research and development of a complete software architecture for Exascale computing. The four major R&D thrusts include: 1) an Exascale lightweight kernel operating system (LXK), based on the Kitten operating system (OS), to manage billion way hardware parallelism, management of faults and power, management of global virtual name space, and other features of future system architectures; 2) a runtime system (HPX-5), co-designed with LXK, based on the ParalleX execution model that supports dynamic resource management and task scheduling; 3) system interfaces for interoperability between the runtime system and both the OS and the runtime system (RIOS); 4) compilation strategies and systems to translate MPI and OpenMP legacy codes to a form that can be run with performance at least as good as a native code implementation. The software architecture development has been driven by performance evaluation of mini-applications developed at Department of Energy Co-Design Centers as well as ancillary in-house developed applications and benchmarks. Research on the ParalleX execution model has been driven by the development of formal semantics and model checkers for the execution model. The major components of the software architecture are: ParalleX execution model – guiding principles for co-design of components of the system stack with advances in locality management and task prioritization through introspection; HPX runtime system – support of application dynamic adaptive resource management, task scheduling, and introspective control policies; LXK operating system – lightweight kernel operating system for order-constant scalability and low/no noise to manage resources; RIOS – a realm of Exascale system operation unique in the XPRESS project that includes the relationship between the new generation of lightweight kernel operating systems and runtime system software; XPI advanced programming model – intermediate form and low-level (readable) programming interface reflecting the ParalleX model, providing a target for source-to- source high level parallel language translation, and supporting early direct programming experimentation and measurement; Performance models & metrics – providing parameters and their mutual sensitivities to guide co-design and quantify operational behavior; Legacy migration – ensuring seamless transition of legacy codes and programming methods to the future generation of ParalleX-based Exascale systems; . Experiments and evaluation – critical to determining degree of effectiveness and likelihood of ultimate success as well as guiding corrective design changes to achieve DOE objectives; and Applications – collaborations with Co-Design Centers and other mission critical codes. The exemplar implementation of the ParalleX execution model, HPX-5, has been released along with five fully ported applications (LULESH, HPCG, SSSP, WAMR, and DASHMM) and multiple microbenchmarks. Using these applications, the system architecture demonstrated dynamic load balancing through active migration, legacy code migration, and data-driven approaches to for graph-like applications. Crucially for Exascale and billion way parallelism, the needs and patterns of irregular applications in HPX-5 were addressed using local flow control rather than relying on globally synchronized data structures.

97 MATHEMATICS AND COMPUTING↗

AMR-Wind: A Performance-Portable, High-Fidelity Flow Solver for Wind Farm Simulations

We present AMR-Wind, a verified and validated high-fidelity computational-fluid-dynamics code for wind farm flows. AMR-Wind is a block-structured, adaptive-mesh, incompressible-flow solver that enables predictive simulations of the atmospheric boundary layer and wind plants. It is a highly scalable code designed for parallel high-performance computing with a specific focus on performance portability for current and future computing architectures, including graphical processing units (GPUs). In this paper, we detail the governing equations, the numerical methods, and the turbine models. Establishing a foundation for the correctness of the code, we present the results of formal verification and validation. The verification studies, which include a novel actuator line test case, indicate that AMR-Wind is spatially and temporally second-order accurate. The validation studies demonstrate that the key physics capabilities implemented in the code, including actuator disk models, actuator line models, turbulence models, and large eddy simulation (LES) models for atmospheric boundary layers, perform well in comparison to reference data from established computational tools and theory. We conclude with a demonstration simulation of a 12-turbine wind farm operating in a turbulent atmospheric boundary layer, detailing computational performance and realistic wake interactions.

17 WIND ENERGY↗

Approximation algorithms in combinatorial scientific computing

We survey recent work on approximation algorithms for computing degree-constrained subgraphs in graphs and their applications in combinatorial scientific computing. The problems we consider include maximization versions of cardinality matching, edge-weighted matching, vertex-weighted matching and edge-weighted $b$ -matching, and minimization versions of weighted edge cover and $b$ -edge cover. Exact algorithms for these problems are impractical for massive graphs with several millions of edges. For each problem we discuss theoretical foundations, the design of several linear or near-linear time approximation algorithms, their implementations on serial and parallel computers, and applications. Our focus is on practical algorithms that yield good performance on modern computer architectures with multiple threads and interconnected processors. We also include information about the software available for these problems.

Mathematics↗

Large-scale parallel multibody dynamics with frictional contact on the GPU.

In the context of simulating the frictional contact dynamics of large systems of rigid bodies, this paper reviews a novel method for solving large cone complementarity problems by means of a fixed-point iteration algorithm. The method is an extension of the Gauss-Seidel and Gauss-Jacobi methods with over-relaxation for symmetric convex linear complementarity problems. Convergent under fairly standard assumptions, the method is implemented in a parallel framework by using a single instruction multiple data computation paradigm promoted by the Compute Unified Device Architecture library for graphical processing unit programming. The framework supports the simulation of problems with more than one million bodies in contact. Simulation thus becomes a viable tool for investigating the dynamics of complex systems such as ground vehicles running on sand, powder composites, and granular material flow.

99 GENERAL AND MISCELLANEOUS↗

Optimizing High Performance Markov Clustering for Pre-Exascale Architectures

HipMCL is a high-performance distributed memory implementation of the popular Markov Cluster Algorithm (MCL) and can cluster large-scale networks within hours using a few thousand CPU-equipped nodes. It relies on sparse matrix computations and heavily makes use of the sparse matrix-sparse matrix multiplication kernel (SpGEMM). The existing parallel algorithms in HipMCL are not scalable to Exascale architectures, both due to their communication costs dominating the runtime at large concurrencies and also due to their inability to take advantage of accelerators that are increasingly popular. In this work, we systematically remove scalability and performance bottlenecks of HipMCL. We enable GPUs by performing the expensive expansion phase of the MCL algorithm on GPU. Additionally, we propose a CPU-GPU joint distributed SpGEMM algorithm called pipelined Sparse SUMMA and integrate a probabilistic memory requirement estimator that is fast and accurate. Furthermore, we develop a new merging algorithm for the incremental processing of partial results produced by the GPUs, which improves the overlap efficiency and the peak memory usage. We also integrate a recent and faster algorithm for performing SpGEMM on CPUs. We validate our new algorithms and optimizations with extensive evaluations. With the enabling of the GPUs and integration of new algorithms, HipMCL is up to 12.4x faster, being able to cluster a network with 70 million proteins and 68 billion connections just under 15 minutes using 1024 nodes of ORNL's Summit supercomputer.

97 MATHEMATICS AND COMPUTING↗

Parallel Agent-Based Simulations on Clusters of GPUs and Multi-Core Processors

An effective latency-hiding mechanism is presented in the parallelization of agent-based model simulations (ABMS) with millions of agents. The mechanism is designed to accommodate the hierarchical organization as well as heterogeneity of current state-of-the-art parallel computing platforms. We use it to explore the computation vs. communication trade-off continuum available with the deep computational and memory hierarchies of extant platforms and present a novel analytical model of the tradeoff. We describe our implementation and report preliminary performance results on two distinct parallel platforms suitable for ABMS: CUDA threads on multiple, networked graphical processing units (GPUs), and pthreads on multi-core processors. Message Passing Interface (MPI) is used for inter-GPU as well as inter-socket communication on a cluster of multiple GPUs and multi-core processors. Results indicate the benefits of our latency-hiding scheme, delivering as much as over 100-fold improvement in runtime for certain benchmark ABMS application scenarios with several million agents. This speed improvement is obtained on our system that is already two to three orders of magnitude faster on one GPU than an equivalent CPU-based execution in a popular simulator in Java. Thus, the overall execution of our current work is over four orders of magnitude faster when executed on multiple GPUs.

97 MATHEMATICS AND COMPUTING↗

A Framework for Lattice QCD Calculations on GPUs

Efficient calculation of the greatest common divisor (GCD) for big integers each whose number of bits is greater than or equal to 1024 has drawn a considerable amount of attention because it can be used to detect a weakness of the RSA security infrastructure. This paper presents a parallel binary GCD algorithm and its implementation for big integers on the Intel Xeon Phi coprocessor. This algorithm is capable of computing GCDs efficiently on many pairs of big integers in parallel by utilizing all cores on a Xeon Phi coprocessor as well as taking advantage of all vector processing units of the coprocessor to speed up critical integer operations within the algorithm. Using 240 threads on a Xeon Phi coprocessor to carry out GCD calculations for a large amount of 2048-bit integers, the implementation achieves the speedup of 30 times over a sequential binary GCD algorithm implementation on a single CPU core, and it delivers twice amount of performance in comparison to the same sequential binary GCD implementation running on 240 threads of the Xeon Phi.

Winter, F. T.↗

A Monte Carlo neutron transport code for eigenvalue calculations on a dual-GPU system and CUDA environment

Monte Carlo (MC) method is able to accurately calculate eigenvalues in reactor analysis. Its lengthy computation time can be reduced by general-purpose computing on Graphics Processing Units (GPU), one of the latest parallel computing techniques under development. The method of porting a regular transport code to GPU is usually very straightforward due to the 'embarrassingly parallel' nature of MC code. However, the situation becomes different for eigenvalue calculation in that it will be performed on a generation-by-generation basis and the thread coordination should be explicitly taken care of. This paper presents our effort to develop such a GPU-based MC code in Compute Unified Device Architecture (CUDA) environment. The code is able to perform eigenvalue calculation under simple geometries on a multi-GPU system. The specifics of algorithm design, including thread organization and memory management were described in detail. The original CPU version of the code was tested on an Intel Xeon X5660 2.8 GHz CPU, and the adapted GPU version was tested on NVIDIA Tesla M2090 GPUs. Double-precision floating point format was used throughout the calculation. The result showed that a speedup of 7.0 and 33.3 were obtained for a bare spherical core and a binary slab system respectively. The speedup factor was further increased by a factor of {approx}2 on a dual GPU system. The upper limit of device-level parallelism was analyzed, and a possible method to enhance the thread-level parallelism was proposed. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Facilitating Co-Design for Extreme-Scale Systems Through Lightweight Simulation

This work focuses on tools for investigating algorithm performance at extreme scale with millions of concurrent threads and for evaluating the impact of future architecture choices to facilitate the co-design of high-performance computing (HPC) architectures and applications. The approach focuses on lightweight simulation of extreme-scale HPC systems with the needed amount of accuracy. The prototype presented in this paper is able to provide this capability using a parallel discrete event simulation (PDES), such that a Message Passing Interface (MPI) application can be executed at extreme scale, and its performance properties can be evaluated. The results of an initial prototype are encouraging as a simple 'hello world' MPI program could be scaled up to 1,048,576 virtual MPI processes on a four-node cluster, and the performance properties of two MPI programs could be evaluated at up to 16,384 virtual MPI processes on the same system.

97 MATHEMATICS AND COMPUTING↗

ParaText : scalable text modeling and analysis.

Automated processing, modeling, and analysis of unstructured text (news documents, web content, journal articles, etc.) is a key task in many data analysis and decision making applications. As data sizes grow, scalability is essential for deep analysis. In many cases, documents are modeled as term or feature vectors and latent semantic analysis (LSA) is used to model latent, or hidden, relationships between documents and terms appearing in those documents. LSA supplies conceptual organization and analysis of document collections by modeling high-dimension feature vectors in many fewer dimensions. While past work on the scalability of LSA modeling has focused on the SVD, the goal of our work is to investigate the use of distributed memory architectures for the entire text analysis process, from data ingestion to semantic modeling and analysis. ParaText is a set of software components for distributed processing, modeling, and analysis of unstructured text. The ParaText source code is available under a BSD license, as an integral part of the Titan toolkit. ParaText components are chained-together into data-parallel pipelines that are replicated across processes on distributed-memory architectures. Individual components can be replaced or rewired to explore different computational strategies and implement new functionality. ParaText functionality can be embedded in applications on any platform using the native C++ API, Python, or Java. The ParaText MPI Process provides a 'generic' text analysis pipeline in a command-line executable that can be used for many serial and parallel analysis tasks. ParaText can also be deployed as a web service accessible via a RESTful (HTTP) API. In the web service configuration, any client can access the functionality provided by ParaText using commodity protocols ... from standard web browsers to custom clients written in any language.

99 GENERAL AND MISCELLANEOUS↗

Proactive Fault Tolerance Using Preemptive Migration

Proactive fault tolerance (FT) in high-performance computing is a concept that prevents compute node failures from impacting running parallel applications by preemptively migrating application parts away from nodes that are about to fail. This paper provides a foundation for proactive FT by defining its architecture and classifying implementation options. This paper further relates prior work to the presented architecture and classification, and discusses the challenges ahead for needed supporting technologies.

97 MATHEMATICS AND COMPUTING↗

Optimization of a parallel hole collimator/CdZnTe gamma-camera architecture for scintimammography

Purpose: Small field-of-view CdZnTe (CZT) gamma cameras are increasingly studied for breast lesion detection to complement mammography or ultrasonographic findings. However, in classical collimation configurations, they remain limited by the trade-off between spatial resolution and sensitivity. The HiSens architecture was proposed to overcome these limitations. Using an accurate 3D localization of the interactions inside the detector, this architecture leads to a gain in sensitivity without loss in spatial resolution. In this article, the relevance of the HiSens architecture for planar scintimammography is studied. Methods: A detective quantum efficiency (DQE) computation method is developed and used to optimize the dimensioning of a parallel hole collimator dedicated to scintimammography. Based on the DQE curves, the impact of the collimator-to-detector distance is studied. Two algorithms are proposed to combine data acquired with different collimator-to-detector distances. Results: It is shown that CZT detector virtual pixelization increases system sensitivity by 3.3 while preserving a standard LEHR spatial resolution. The introduction of a gap between the CZT detector and the collimator is useful to modulate the DQE curve shape. The combination of data acquired using different gaps in the image formation process leads to enhanced restoration of the frequency content of the images, resulting in image contrast and spatial resolution improvements. Conclusions: Acquisition duration or injected activity could be markedly reduced if the HiSens architecture with an appropriate collimator-detector gap were used.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

The Kokkos OpenMPTarget Backend: Implementation and Lessons Learned

As the supercomputing landscape diversifies, solutions such as Kokkos to write vendor agnostic applications and libraries have risen in popularity. Kokkos provides a programming model designed for performance portability, which allows developers to write a single source implementation that can run efficiently on various architectures. At its heart, Kokkos maps parallel algorithms to architecture and vendor specific backends written in lower level programming models such as CUDA and HIP. Another approach to writing vendor agnostic parallel code is using OpenMP’s directives based approach, which lets developers annotate code to express parallelism. It is implemented at the compiler level and is supported by all major high performance computing vendors, as well as the primary Open Source toolchains GNU and LLVM. Since its inception, Kokkos has used OpenMP to parallelize on CPU architectures. In this paper, we explore leveraging OpenMP for a GPU backend and discuss the challenges we encountered when mapping the Kokkos APIs and semantics to OpenMP target constructs. As an exemplar workload we chose a simple conjugate gradient solver for sparse matrices. We find that performance on NVIDIA and AMD GPUs varies widely based on details of the implementation strategy and the chosen compiler. Furthermore, the performance of the OpenMP implementations decreases with increasing complexity of the investigated algorithms.

Gayatri, Rahulkumar↗

Speeding up particle track reconstruction using a parallel Kalman filter algorithm

One of the most computationally difficult problems expected for the High-Luminosity Large Hadron Collider (HL-LHC) is determining the trajectory of charged particles during event reconstruction. Algorithms used at the LHC today rely on Kalman filtering, which builds physical trajectories incrementally while incorporating material effects and error estimation. Recognizing the need for faster computational throughput, we have adapted Kalman-filter-based methods for highly parallel, many-core SIMD architectures that are now prevalent in high-performance hardware. In this paper, we discuss the design and performance of the improved tracking algorithm, referred to as mkFit. A key piece of the algorithm is the Matriplex library, containing dedicated code to optimally vectorize operations on small matrices. The physics performance of the mkFit algorithm is comparable to the nominal CMS tracking algorithm when reconstructing tracks from simulated proton-proton collisions within the CMS detector. We study the scaling of the algorithm as a function of the parallel resources utilized and find large speedups both from vectorization and multi-threading. mkFit achieves a speedup of a factor of 6 compared to the nominal algorithm when run in a single-threaded application within the CMS software framework.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Exploring Heterogeneous Multicore Architectures for Advanced Embedded Uncertainty Quantification

We explore rearrangements of classical uncertainty quantification methods with the aim of achieving higher aggregate performance for uncertainty quantification calculations on emerging multicore and many core architectures. We show a rearrangement of the stochastic Galerkin method leads to improved performance and scalability on several computational architectures whereby uncertainty information is propagated at the lowest levels of the simulation code improving memory access patterns, exposing new dimensions of fine grained parallelism, and reducing communication. We also develop a general framework for implementing such rearrangements for a diverse set of uncertainty quantification algorithms as well as computational simulation codes to which they are applied.

97 MATHEMATICS AND COMPUTING↗