arXiv subjects

Boyana Norris

Publications and source records attributed to Boyana Norris.

23 records · Page 2Linked to original sources

arXiv2017

A Comparison of Parallel Graph Processing Implementations

The rapidly growing number of large network analysis problems has led to the emergence of many parallel and distributed graph processing systems---one survey in 2014 identified over 80. Since then, the landscape has evolved; some packages have become inactive while more are being developed. Determining the best approach for a given problem is infeasible for most developers. To enable easy, rigorous, and repeatable comparison of the capabilities of such systems, we present an approach and associated software for analyzing the performance and scalability of parallel, open-source graph libraries. We demonstrate our approach on five graph processing packages: GraphMat, the Graph500, the Graph Algorithm Platform Benchmark Suite, GraphBIG, and PowerGraph using synthetic and real-world datasets. We examine previously overlooked aspects of parallel graph processing performance such as phases of execution and energy usage for three algorithms: breadth first search, single source shortest paths, and PageRank and compare our results to Graphalytics.

cs.PF↗

arXiv2015

A Roofline Visualization Framework

The Roofline Model and its derivatives provide an intuitive representation of the best achievable performance on a given architecture. The Roofline Toolkit project is a collaboration among researchers at Argonne National Laboratory, Lawrence Berkeley National Laboratory, and the University of Oregon and consists of three main parts: hardware characterization, software characterization, and data manipulation and visualization interface. These components address the different aspects of performance data acquisition and manipulation required for performance analysis, modeling and optimization of codes on existing and emerging architectures. In this paper we introduce an initial implementation of the third component, a system for visualizing roofline charts and managing roofline performance analysis data. We discuss the implementation and rationale for the integration of the roofline visualization system into the Eclipse IDE. An overview of our continuing efforts and goals in the development of this project is provided.

cs.PF↗

arXiv2014

Lighthouse: A User-Centered Web Service for Linear Algebra Software

Various fields of science and engineering rely on linear algebra for large scale data analysis, modeling and simulation, machine learning, and other applied problems. Linear algebra computations often dominate the execution time of such applications. Meanwhile, experts in these domains typically lack the training or time required to develop efficient, high-performance implementations of linear algebra algorithms. In the Lighthouse project, we enable developers with varied backgrounds to readily discover and effectively apply the best available numerical software for their problems. We have developed a search-based expert system that combines expert knowledge, machine learningbased classification of existing numerical software collections, and automated code generation and optimization. Lighthouse provides a novel software engineering environment aimed at maximizing both developer productivity and application performance for dense and sparse linear algebra computations.

cs.MS↗

arXiv2013

Software Autotuning for Sustainable Performance Portability

Scientific software applications are increasingly developed by large interdiscplinary teams operating on functional modules organized around a common software framework, which is capable of integrating new functional capabilities without modifying the core of the framework. In such environment, software correctness and modularity take precedence at the expense of code performance, which is an important concern during execution on supercomputing facilities, where the allocation of core-hours is a valuable resource. To alleviate the performance problems, we propose automated performance tuning (autotuning) of software to extract the maximum performance on a given hardware platform and to enable performance portability across heterogeneous hardware platforms. The resulting code remains generic without committing to a particular software stack and yet is compile-time specializable for maximal sustained performance.

cs.PF↗

arXiv2012

Reliable Generation of High-Performance Matrix Algebra

Scientific programmers often turn to vendor-tuned Basic Linear Algebra Subprograms (BLAS) to obtain portable high performance. However, many numerical algorithms require several BLAS calls in sequence, and those successive calls result in suboptimal performance. The entire sequence needs to be optimized in concert. Instead of vendor-tuned BLAS, a programmer could start with source code in Fortran or C (e.g., based on the Netlib BLAS) and use a state-of-the-art optimizing compiler. However, our experiments show that optimizing compilers often attain only one-quarter the performance of hand-optimized code. In this paper we present a domain-specific compiler for matrix algebra, the Build to Order BLAS (BTO), that reliably achieves high performance using a scalable search algorithm for choosing the best combination of loop fusion, array contraction, and multithreading for data parallelism. The BTO compiler generates code that is between 16% slower and 39% faster than hand-optimized code.

cs.MS↗