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 433 records · Page 24

Scalable Static and Dynamic Community Detection Using Grappolo

Graph clustering, popularly known as community detection, is a fundamental kernel for several applications of relevance to the Defense Advanced Research Projects Agency’s (DARPA) Hierarchical Identify Verify Exploit (HIVE) Pro- gram. Clusters or communities represent natural divisions within a network that are densely connected within a cluster and sparsely connected to the rest of the network. The need to compute clustering on large scale data necessitates the development of efficient algorithms that can exploit modern architectures that are fundamentally parallel in nature. How- ever, due to their irregular and inherently sequential nature, many of the current algorithms for community detection are challenging to parallelize. In response to the HIVE Graph Challenge, we present several parallelization heuristics for fast community detection using the Louvain method as the serial template. We implement all the heuristics in a software library called Grappolo. Using the inputs from the HIVE Challenge, we demonstrate superior performance and high quality solutions based on four parallelization heuristics. We use Grappolo on static graphs as the first step towards community detection on streaming graphs.

Community Detection↗

Extracting SIMD Parallelism from Recursive Task-Parallel Programs

The pursuit of computational efficiency has led to the proliferation of throughput-oriented hardware, from GPUs to increasingly wide vector units on commodity processors and accelerators. This hardware is designed to efficiently execute data-parallel computations in a vectorized manner. However, many algorithms are more naturally expressed as divide-and-conquer, recursive, task-parallel computations. In the absence of data parallelism, it seems that such algorithms are not well suited to throughput-oriented architectures. This paper presents a set of novel code transformations that expose the data parallelism latent in recursive, task- parallel programs. These transformations facilitate straightforward vectorization of task-parallel programs on commodity hardware. We also present scheduling policies that maintain high utilization of vector resources while limiting space usage. Across several task-parallel benchmarks, we demonstrate both efficient vector resource utilization and substantial speedup on chips using Intel’s SSE4.2 vector units, as well as accelerators using Intel’s AVX512 units. We then show through rigorous sampling that, in practice, our vectorization techniques are effective for a much larger class of programs.

Recursive Programs, task parallelism, vectorizatio↗

A Programming Model for Massive Data Parallelism with Data Dependencies

Accelerating processors can often be more cost and energy effective for a wide range of data-parallel computing problems than general-purpose processors. For graphics processor units (GPUs), this is particularly the case when program development is aided by environments such as NVIDIA s Compute Unified Device Architecture (CUDA), which dramatically reduces the gap between domain-specific architectures and general purpose programming. Nonetheless, general-purpose GPU (GPGPU) programming remains subject to several restrictions. Most significantly, the separation of host (CPU) and accelerator (GPU) address spaces requires explicit management of GPU memory resources, especially for massive data parallelism that well exceeds the memory capacity of GPUs. One solution to this problem is to transfer data between the GPU and host memories frequently. In this work, we investigate another approach. We run massively data-parallel applications on GPU clusters. We further propose a programming model for massive data parallelism with data dependencies for this scenario. Experience from micro benchmarks and real-world applications shows that our model provides not only ease of programming but also significant performance gains.

43 PARTICLE ACCELERATORS↗

High-Performance Computer Modeling of the Cosmos-Iridium Collision

This paper describes the application of a new, integrated modeling and simulation framework, encompassing the space situational awareness (SSA) enterprise, to the recent Cosmos-Iridium collision. This framework is based on a flexible, scalable architecture to enable efficient simulation of the current SSA enterprise, and to accommodate future advancements in SSA systems. In particular, the code is designed to take advantage of massively parallel, high-performance computer systems available, for example, at Lawrence Livermore National Laboratory. We will describe the application of this framework to the recent collision of the Cosmos and Iridium satellites, including (1) detailed hydrodynamic modeling of the satellite collision and resulting debris generation, (2) orbital propagation of the simulated debris and analysis of the increased risk to other satellites (3) calculation of the radar and optical signatures of the simulated debris and modeling of debris detection with space surveillance radar and optical systems (4) determination of simulated debris orbits from modeled space surveillance observations and analysis of the resulting orbital accuracy, (5) comparison of these modeling and simulation results with Space Surveillance Network observations. We will also discuss the use of this integrated modeling and simulation framework to analyze the risks and consequences of future satellite collisions and to assess strategies for mitigating or avoiding future incidents, including the addition of new sensor systems, used in conjunction with the Space Surveillance Network, for improving space situational awareness.

42 ENGINEERING↗

A pervasive parallel framework for visualization: final report for FWP 10-014707

We are on the threshold of a transformative change in the basic architecture of highperformance computing. The use of accelerator processors, characterized by large core counts, shared but asymmetrical memory, and heavy thread loading, is quickly becoming the norm in high performance computing. These accelerators represent significant challenges in updating our existing base of software. An intrinsic problem with this transition is a fundamental programming shift from message passing processes to much more fine thread scheduling with memory sharing. Another problem is the lack of stability in accelerator implementation; processor and compiler technology is currently changing rapidly. This report documents the results of our three-year ASCR project to address these challenges. Our project includes the development of the Dax toolkit, which contains the beginnings of new algorithms for a new generation of computers and the underlying infrastructure to rapidly prototype and build further algorithms as necessary.

97 MATHEMATICS AND COMPUTING↗

Parallel peak pruning for scalable SMP contour tree computation

As data sets grow to exascale, automated data analysis and visualisation are increasingly important, to intermediate human understanding and to reduce demands on disk storage via in situ analysis. Trends in architecture of high performance computing systems necessitate analysis algorithms to make effective use of combinations of massively multicore and distributed systems. One of the principal analytic tools is the contour tree, which analyses relationships between contours to identify features of more than local importance. Unfortunately, the predominant algorithms for computing the contour tree are explicitly serial, and founded on serial metaphors, which has limited the scalability of this form of analysis. While there is some work on distributed contour tree computation, and separately on hybrid GPU-CPU computation, there is no efficient algorithm with strong formal guarantees on performance allied with fast practical performance. Here in this paper, we report the first shared SMP algorithm for fully parallel contour tree computation, withfor-mal guarantees of O(lgnlgt) parallel steps and O(n lgn) work, and implementations with up to 10x parallel speed up in OpenMP and up to 50x speed up in NVIDIA Thrust.

97 MATHEMATICS AND COMPUTING↗

An efficient MPI/OpenMP parallelization of the Hartree–Fock–Roothaan method for the first generation of Intel® Xeon Phi™ processor architecture

The Hartree-Fock (HF) method in the quantum chemistry package GAMESS represents one of the most irregular algorithms in computation today. Major steps in the calculation are the irregular computation of electron repulsion integrals (ERIs) and the building of the Fock matrix. These are the central components of the main Self Consistent Field (SCF) loop, the key hotspot in Electronic Structure (ES) codes. By threading the MPI ranks in the official release of the GAMESS code, we not only speed up the main SCF loop (4x to 6x for large systems), but also achieve a significant (>2x) reduction in the overall memory footprint. These improvements are a direct consequence of memory access optimizations within the MPI ranks. We benchmark our implementation against the official release of the GAMESS code on the Intel R Xeon PhiTM supercomputer. Here, scaling numbers are reported on up to 7,680 cores on Intel Xeon Phi coprocessors.

97 MATHEMATICS AND COMPUTING↗

Optimal Operation of a Hybrid Hydraulic Electric Architecture (HHEA) for Off-Road Vehicles Over Discrete Operating Decisions

Many off-highway machines including construction and agriculture equipment use hydraulics for power transmission and throttling as a means for control. Trends towards better efficiency and electrification have led to the creation of a novel Hybrid Hydraulic-Electric Architecture (HHEA) which could significantly reduce energy consumption and maintain control performance, even in machines that are too large to be directly electrified. This is achieved by using a set of common pressure rails to transmit the majority of power via power dense hydraulics and modulating the power with small electric motor-drives to achieve precise control. This paper proposes a computationally efficient method for computing the optimal sequence of pressure rail selections for the HHEA over finite drive cycles. This is useful for fairly comparing the novel architecture’s energy performance to existing architectures and for use in iterative optimal design of the architecture. The optimal control technique makes use of a static model of the architecture and losses. Constraints are enforced such that the drive cycle is repeatable. The constrained optimal operation is solved using a Lagrange multiplier technique that transforms the optimization into a small set of sub-problems by considering combinations of active constraints. Each of these sub-problems can be solved efficiently because loss calculations for all time steps can be computed in parallel. A case study of an off-road construction machine demon-strates that the HHEA reduces energy consumption by 2/3 compared to the baseline load sensing architecture.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Summer Internship Report: ARA2 Benchmarking

Over the past decade, the RISC-V Instruction Set Architecture (ISA) has emerged as a significant player in both academic and industrial processor design due to its open-source nature, modular extension system, and versatility across domains ranging from microcontrollers to high-performance computing (HPC). One of its most important recent advancements is the RISC-V Vector Extension (RVV), which enables explicit data-level parallelism through vector registers and vectorized instructions. Unlike traditional SIMD (Single Instruction, Multiple Data) architectures that fix vector lengths at design time, RVV uses the concept of VLEN (vector register length) as a hardware-independent parameter and allows software to adapt dynamically to the available vector width. This flexible approach ensures portability across implementations while enabling scalable performance. The ARA2 core is a parameterizable RISC-V vector processor developed at the Integrated Systems Lab at ETH Zürich and the University of Bologna. Designed as a tightly-coupled accelerator to a scalar RISC-V core, ARA2 implements the RVV 1.0 specification and offers tunable architectural parameters such as the number of vector lanes, VLEN, and cache sizes.

97 MATHEMATICS AND COMPUTING↗

Dense LU Factorization on Multicore Supercomputer Nodes

Dense LU factorization is a prominent benchmark used to rank the performance of supercomputers. Many implementations, including the reference code HPL, use block-cyclic distributions of matrix blocks onto a two-dimensional process grid. The process grid dimensions drive a trade-off between communication and computation and are architecture- and implementation-sensitive. We show how the critical panel factorization steps can be made less communication-bound by overlapping asynchronous collectives for pivot identification and exchange with the computation of rank-k updates. By shifting this trade-off, a modified block-cyclic distribution can beneficially exploit more available parallelism on the critical path, and reduce panel factorization's memory hierarchy contention on now-ubiquitous multi-core architectures. The missed parallelism in traditional block-cyclic distributions arises because active panel factorization, triangular solves, and subsequent broadcasts are spread over single process columns or rows (respectively) of the process grid. Increasing one dimension of the process grid decreases the number of distinct processes in the other dimension. To increase parallelism in both dimensions, periodic 'rotation' is applied to the process grid to recover the row-parallelism lost by a tall process grid. During active panel factorization, rank-1 updates stream through memory with minimal reuse. In a column-major process grid, the performance of this access pattern degrades as too many streaming processors contend for access to memory. A block-cyclic mapping in the more popular row-major order does not encounter this problem, but consequently sacrifices node and network locality in the critical pivoting steps. We introduce 'striding' to vary between the two extremes of row- and column-major process grids. As a test-bed for further mapping experiments, we describe a dense LU implementation that allows a block distribution to be defined as a general function of block to processor. Other mappings can be tested with only small, local changes to the code.

99 GENERAL AND MISCELLANEOUS↗

Parallel-in-Time Multigrid Methods for Hyperbolic Problems, with a Focus on the Shallow Water Equations

The coming massive parallelism of exascale computing presents a pressing challenge for the many DOE simulations of time-dependent partial differential equations, which typically use traditional sequential time stepping methods. Since this traditional approach is inherently serial, it presents a sequential bottleneck when moving to exascale computing, because future performance gains will come through greater concurrency, not faster clock speeds. Thus, the goal of this work is to research parallelism in time, i.e., methods that compute multiple time values simultaneously, not sequentially. The focus will be on hyperbolic problems of programmatic interest to DOE, with the goal of enabling scalable simulations of time-dependent hyperbolic problems on future architectures. The difficulty lies in the fact that hyperbolic problems are well-known to be difficult for parallel-in-time methods, with the most common parallel-in-time method, Parareal, diverging in many cases. Here, the chosen methodology for scalably solving these hyperbolic space-time equations is multigrid-reductionin-time (MGRIT), because multigrid (when it works) is a powerful, optimal, and scalable solver for discretized PDEs. To further research in this area, this project will explore a model hyperbolic problem (the shallow water equations) in the context of recent advances (e.g., by Wingate and Haut) in constructing improved coarse time-grid time-propagators by using the slow asymptotic structure of the equations. In particular, we will investigate if such coarse time-propagators offer significant advantages in an MGRIT setting, and explore any broader insights gained into parallel-in-time for hyperbolic problems.

97 MATHEMATICS AND COMPUTING↗

Massively parallel implementations of coupled-cluster methods for electron spin resonance spectra. I. Isotropic hyperfine coupling tensors in large radicals

Coupled cluster (CC) methods provide highly accurate predictions of molecular properties, but their high computational cost has precluded their routine application to large systems. Fortunately, recent computational developments in the ACES III program by the Bartlett group [the OED/ERD atomic integral package, the super instruction processor, and the super instruction architecture language] permit overcoming that limitation by providing a framework for massively parallel CC implementations. In that scheme, we are further extending those parallel CC efforts to systematically predict the three main electron spin resonance (ESR) tensors (A-, g-, and D-tensors) to be reported in a series of papers. In this paper inaugurating that series, we report our new ACES III parallel capabilities that calculate isotropic hyperfine coupling constants in 38 neutral, cationic, and anionic radicals that include the {sup 11}B, {sup 17}O, {sup 9}Be, {sup 19}F, {sup 1}H, {sup 13}C, {sup 35}Cl, {sup 33}S,{sup 14}N, {sup 31}P, and {sup 67}Zn nuclei. Present parallel calculations are conducted at the Hartree-Fock (HF), second-order many-body perturbation theory [MBPT(2)], CC singles and doubles (CCSD), and CCSD with perturbative triples [CCSD(T)] levels using Roos augmented double- and triple-zeta atomic natural orbitals basis sets. HF results consistently overestimate isotropic hyperfine coupling constants. However, inclusion of electron correlation effects in the simplest way via MBPT(2) provides significant improvements in the predictions, but not without occasional failures. In contrast, CCSD results are consistently in very good agreement with experimental results. Inclusion of perturbative triples to CCSD via CCSD(T) leads to small improvements in the predictions, which might not compensate for the extra computational effort at a non-iterative N{sup 7}-scaling in CCSD(T). The importance of these accurate computations of isotropic hyperfine coupling constants to elucidate experimental ESR spectra, to interpret spin-density distributions, and to characterize and identify radical species is illustrated with our results from large organic radicals. Those include species relevant for organic chemistry, petroleum industry, and biochemistry, such as the cyclo-hexyl, 1-adamatyl, and Zn-porphycene anion radicals, inter alia.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Towards Highly Scalable Ab Initio Molecular Dynamics (AIMD) Simulations on the Intel Knights Landing Manycore Processor

The Ab Initio Molecular Dynamics (AIMD) method allows scientists to treat the dynamics of molecular and condensed phase systems while retaining a first-principles-based description of their interactions. This extremely important method has tremendous computational requirements, because the electronic Schr¨odinger equation, approximated using Kohn-Sham Density Functional Theory (DFT), is solved at every time step. With the advent of manycore architectures, application developers have a significant amount of processing power within each compute node that can only be exploited through massive parallelism. A compute intensive application such as AIMD forms a good candidate to leverage this processing power. In this paper, we focus on adding thread level parallelism to the plane wave DFT methodology implemented in NWChem. Through a careful optimization of tall-skinny matrix products, which are at the heart of the Lagrange multiplier and nonlocal pseudopotential kernels, as well as 3D FFTs, our OpenMP implementation delivers excellent strong scaling on the latest Intel Knights Landing (KNL) processor. We assess the efficiency of our Lagrange multiplier kernels by building a Roofline model of the platform, and verify that our implementation is close to the roofline for various problem sizes. Finally, we present strong scaling results on the complete AIMD simulation for a 64 water molecules test case, that scales up to all 68 cores of the Knights Landing processor.

Environmental Molecular Sciences Laboratory↗

Efficient Execution of Recursive Programs on Commodity Vector Hardware

The pursuit of computational efficiency has led to the proliferation of throughput-oriented hardware, from GPUs to increasingly-wide vector units on commodity processors and accelerators. This hardware is designed to efficiently execute data-parallel computations in a vectorized manner. However, many algorithms are more naturally expressed as divide-and-conquer, recursive, task-parallel computations; in the absence of data parallelism, it seems that such algorithms are not well-suited to throughput-oriented architectures. This paper presents a set of novel code transformations that expose the data-parallelism latent in recursive, task-parallel programs. These transformations facilitate straightforward vectorization of task-parallel programs on commodity hardware. We also present scheduling policies that maintain high utilization of vector resources while limiting space usage. Across several task-parallel benchmarks, we demonstrate both efficient vector resource utilization and substantial speedup on chips using Intel's SSE4.2 vector units as well as accelerators using Intel's AVX512 units.

resursion↗

Moment representation in the lattice Boltzmann method on massively parallel hardware

The widely-used lattice Boltzmann method (LBM) for computational fluid dynamics is highly scalable, but also significantly memory bandwidth-bound on current architectures. This paper presents a new regularized LBM implementation that reduces the memory footprint by only storing macroscopic, moment-based data. We show that the amount of data that must be stored in memory during a simulation is reduced by up to 47%. We also present a technique for cache-aware data re-utilization and show that optimizing cache utilization to limit data motion results in a similar improvement in time to solution. These new algorithms are implemented in the hemodynamics solver HARVEY and demonstrated using both idealized and realistic biological geometries. We develop a performance model for the moment representation algorithm and evaluate the performance on Summit.

Vardhan, Madhurima↗

Parallel Multigrid in Time and Space for Extreme-Scale Computational Science: Chaotic and Hyperbolic Problems

The coming massive parallelism of exascale computing presents a pressing challenge for the many DOE simulations of time-dependent partial differential equations (PDEs), which typically use traditional sequential time stepping methods. Since this traditional approach is inherently serial, it presents a sequential bottleneck when moving to exascale computing, because future performance gains will come through greater concurrency, not faster clock speeds. Thus, the goal of this work is to research parallelism in time, i.e., methods that compute multiple time values simultaneously, not sequentially. The focus will be on hyperbolic and chaotic problems of interest to DOE, with the goal of enabling scalable simulations of time-dependent hyperbolic and chaotic problems on future architectures. The chosen methodology for solving these problems parallel-in-time is multigrid, because multigrid (when it works) is a powerful, optimal, and scalable solver for discretized PDEs. Multigrid is already commonly used in many DOE simulations for scalably and optimally solving space-only PDE problems. The areas of hyperbolic and chaotic problems are chosen because of their relevance to problems of programmatic interest to DOE. However, these problems are also well-known to be difficult for parallelin-time methods, with the most common method, parareal, diverging in many cases. The current stateof-the-art for parallel-in-time at LLNL is the multigrid reduction in time (MGRIT) XBraid package, which also struggles for such problems, while still showing some improvement over parareal. In summary, new methods are needed for an efficient parallel-in-time scheme for hyperbolic and chaotic problems, and this work shall research promising new multigrid methods in this area. In particular, we take inspiration from the Least Squares Shadowing (LSS by Wang) approach for solving chaotic problems. Here, an optimization approach is able to find “well-conditioned” shadow trajectories/solutions to the original “ill-conditioned” chaotic problem. Thus, the new multigrid methods researched here also arise in an optimization context.

97 MATHEMATICS AND COMPUTING↗

Power and Performance Trade-offs for Space Time Adaptive Processing

Computational efficiency – performance relative to power or energy – is one of the most important concerns when designing RADAR processing systems. This paper analyzes power and performance trade-offs for a typical Space Time Adaptive Processing (STAP) application. We study STAP implementations for CUDA and OpenMP on two computationally efficient architectures, Intel Haswell Core I7-4770TE and NVIDIA Kayla with a GK208 GPU. We analyze the power and performance of STAP’s computationally intensive kernels across the two hardware testbeds. We also show the impact and trade-offs of GPU optimization techniques. We show that data parallelism can be exploited for efficient implementation on the Haswell CPU architecture. The GPU architecture is able to process large size data sets without increase in power requirement. The use of shared memory has a significant impact on the power requirement for the GPU. A balance between the use of shared memory and main memory access leads to an improved performance in a typical STAP application.

Gawande, Nitin A.↗