Search NASA⌕ Search

SEARCH · Search NASA

Results for “Multicore”

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 91 records · Page 5

High-Performance Sparse Matrix-Matrix Products on Intel KNL and Multicore Architectures

Sparse matrix-matrix multiplication (SpGEMM) is a computational primitive that is vastly used in areas ranging from traditional numerical applications to recent big data analysis and machine learning. While many SpGEMM algorithms have been proposed, hardware specific optimizations for multi- and many-core processors are lacking and a detailed analysis of their performance under various use cases and matrices is not available. We firstly identify and mitigate multiple bottlenecks with memory management and thread scheduling on Intel Xeon Phi (Knights Landing or KNL). Specifically targeting multi- and many-core processors, we develop a hash-table-based algorithm and optimize a heap-based shared-memory SpGEMM algorithm. We investigate their performance together with other publicly available codes. Different from the literature, our evaluation also includes use cases that are representative of real graph algorithms, such as multi-source breadth-first search or triangle counting. Our hash-table and heap-based algorithms are showing significant speedups from libraries in the majority of the cases while different algorithms dominate the other scenarios with different matrix size, sparsity, compression factor and operation type. We wrap up in-depth evaluation results and make a recipe to give the best SpGEMM algorithm for target scenario. In conclusion, a critical finding is that hash-table-based SpGEMM gets a significant performance boost if the nonzeros are not required to be sorted within each row of the output matrix.

97 MATHEMATICS AND COMPUTING↗

HBMax: Optimizing Memory Efficiency for Parallel Influence Maximization on Multicore Architectures

The goal of influence maximization is to select k most-influential vertices or seeds in a network, where influence is defined by a given diffusion process. The problem has a number of important applications such as viral marketing, information spread, and epidemic control. Although computing optimal seed set is NP-Hard, due to the submodular nature of the problem efficient approximation algorithms exist. However, even state-of-the-art parallel implementations are limited by a sampling step that incurs large memory footprints. This in turn limits the problem size reach and approximation quality. In this work, we study the memory footprint of the sampling process collecting reverse reachability information in the IMM algorithm over large real-world social networks. We present an adaptive and memory-efficient optimization approach for a state-of-the-art multi-threaded parallel influence maximization algorithm. Our approach,HuffMax, uses a portion of the reverse reachable (RR) sets collected by the algorithm to learn the characteristics of the graph. Then, it compresses the intermediate reverse reachability information with Huffman coding, and queries directly on the compressed data to preserve the memory savings obtained through compression. We also propose an efficient sampling strategy based on the distribution of RR sets, which can further reduce the computation time for typical social networks with long-tail distributions. Considering a NUMA architecture, we scale up our solution on 128-core CPUs and reduce the memory footprint by up to 45.7% with negligible time overhead (or even faster) and without perceivable loss of accuracy.

Chen, Xinyu↗

Optimization of High-order Wave Equations for Multicore CPUs

This is a simple benchmark to gauge the performance of a high-order isotropic wave equation grid. The code is optimized for both SSE and AVX and is parallelized using OpenMP (see Optimization section). Structurally, the benchmark begins, reads a few command-line parameters, allocates and pads the four arrays (current, last, next wave fields, and the spatially varying but isotropic velocity), initializes these arrays, then runs the benchmark proper. The code then benchmarks the naïve, SSE (if supported), and AVX (if supported implementations) by applying the wave equation stencil 100 times and taking the average performance. Boundary conditions are ignored and would nominally be implemented by the user. Thus, the benchmark measures only the performance of the wave equation stencil and not a full simulation. The naïve implementation is a quadruply (z,y,x, radius) nested loop that can handle arbitrarily order wave equations. The optimized (SSE/AVX) implentations are somewhat more complex as they operate on slabs and include a case statement to select an optimized inner loop depending on wave equation order.

Williams, Samuel↗

Auto-tuning Stencil Computations on Multicore and Accelerators

The recent transformation from an environment where gains in computational performance came from increasing clock frequency and other hardware engineering innovations, to an environment where gains are realized through the deployment of ever increasing numbers of modest performance cores has profoundly changed the landscape of scientific application programming. This exponential increase in core count represents both an opportunity and a challenge: access to petascale simulation capabilities and beyond will require that this concurrency be efficiently exploited.

43 PARTICLE ACCELERATORS↗

Optimization of CPAPR for x64 multicore

I report the progress to date of my work on scaling the CPAPR algorithm and necessary supporting code to enable processing large (gigabyte to 100 gigabyte) data sets and benchmarking the same. Where possible, I also report background information possibly of relevance in future modifications of the code. The results include: minor repairs and additions to the TTB library for portability, algorithmic improvements relevant to both serial and multithreaded implementations, algorithmic improvements taking advantage of multithreading hardware, support library additions (binary IO routines) needed for efficiently and reproducibly benchmarking the algorithms. For this optimization work, no large scale data sets are available. Therefore, scalability of data synthesis algorithms is addressed as well.

97 MATHEMATICS AND COMPUTING↗

Multicore Architecture-aware Scientific Applications

Modern high performance systems are becoming increasingly complex and powerful due to advancements in processor and memory architecture. In order to keep up with this increasing complexity, applications have to be augmented with certain capabilities to fully exploit such systems. These may be at the application level, such as static or dynamic adaptations or at the system level, like having strategies in place to override some of the default operating system polices, the main objective being to improve computational performance of the application. The current work proposes two such capabilites with respect to multi-threaded scientific applications, in particular a large scale physics application computing ab-initio nuclear structure. The first involves using a middleware tool to invoke dynamic adaptations in the application, so as to be able to adjust to the changing computational resource availability at run-time. The second involves a strategy for effective placement of data in main memory, to optimize memory access latencies and bandwidth. These capabilties when included were found to have a significant impact on the application performance, resulting in average speedups of as much as two to four times.

97 MATHEMATICS AND COMPUTING↗

Recovery Act: Integrated DC-DC Conversion for Energy-Efficient Multicore Processors

In this project, we have developed the use of thin-film magnetic materials to improve in energy efficiency of digital computing applications by enabling integrated dc-dc power conversion and management with on-chip power inductors. Integrated voltage regulators also enables fine-grained power management, by providing dynamic scaling of the supply voltage in concert with the clock frequency of synchronous logic to throttle power consumption at periods of low computational demand. The voltage converter generates lower output voltages during periods of low computational performance requirements and higher output voltages during periods of high computational performance requirements. Implementation of integrated power conversion requires high-capacity energy storage devices, which are generally not available in traditional semiconductor processes. We achieve this with integration of thin-film magnetic materials into a conventional complementary metal-oxide-semiconductor (CMOS) process for high-quality on-chip power inductors.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Empirical Memory-Access Cost Models in Multicore NUMA Architectures

Data location is of prime importance when scheduling tasks in a non-uniform memory access (NUMA) architecture. The characteristics of the NUMA architecture must be understood so tasks can be scheduled onto processors that are close to the task's data. However, in modern NUMA architectures, such as AMD Magny-Cours and Intel Nehalem, there may be a relatively large number of memory controllers with sockets that are connected in a non-intuitive manner, leading to performance degradation due to uninformed task-scheduling decisions. In this paper, we provide a method for experimentally characterizing memory-access costs for modern NUMA architectures via memory latency and bandwidth microbenchmarks. Using the results of these benchmarks, we propose a memory-access cost model to improve task-scheduling decisions by scheduling tasks near the data they need. Simple task-scheduling experiments using the memory-access cost models validate the use of empirical memory-access cost models to significantly improve program performance.

BENCHMARKS↗

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↗