Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Data Movement”

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 145 records · Page 8

Integrated End-to-end Performance Prediction and Diagnosis for Extreme Scientific Workflows

This report details recent progress for the ASCR funded project “Integrated End-to-end Performance Prediction and Diagnosis for Extreme Scientific Workflows”. We refer to the project as IPPD/2, reflecting the 2017 renewal under expanded scope and partners In IPPD/2, we increased our research scope to include data motion. We are focusing on three major aspects: a) observe how data is generated, distributed, and used; b) analyze how data is (repeatedly) consumed with a focus both on repeated patterns and anomalies; and c) explore how to optimize data motion. This new work on data motion will augment and complement IPPD/2’s research that focused on the computational aspects of tasks. We leverage and extend our existing tools and demonstrate our work on the Belle II workflow suite as well as on workflows from NSLS-II. The highlights of our work are as follows: Provenance for Workflows: Provenance is used to provide information enabling quality control, re-run computational workflows, and reproduce results. IPPD/2 has been building a scalable provenance management system that enables the capture of provenance from the high-level workflow through all relevant system levels in one integrated environment. Leveraging this work, our recent efforts have included using provenance as an enabling technique. Workload characterization: Leveraging provenance and analysis, we characterize data movement within network, storage, and memory over a variety of workloads. This characterization enables an understanding by performance analysts and application developers of the range of behaviors that could be expected. Performance Prediction for Workflows: The goal of modeling distributed workflows is to understand performance bottlenecks and enable more intelligent task scheduling to optimize selected metrics of interest (e.g., task throughput or output data rate). IPPD/2 has utilized both analytical and AI/ML modeling methodologies for performance modeling. Advanced Scheduling and Fault Modeling for Workflows: Scheduling of large-scale scientific workflows on geographically distributed resources is a challenging problem. To improve workflow throughput, we combined novel scheduling algorithms with task predictions from performance modeling and fault modeling. Dynamically Alleviating Bottlenecks in Workflows: Exploiting our provenance, analysis, and modeling efforts, we have explored and developed several techniques for dynamically detecting and alleviating bottlenecks in data movement. In particular, we have spent considerable effort demonstrating our techniques on production-like workflow configurations.

97 MATHEMATICS AND COMPUTING↗

Compiler-Directed Transformation for Higher-Order Stencils

As the cost of data movement increasingly dominates performance, developers of finite-volume and finite-difference solutions for partial differential equations (PDEs) are exploring novel higher-order stencils that increase numerical accuracy and computational intensity. This paper describes a new compiler reordering transformation applied to stencil operators that performs partial sums in buffers, and reuses the partial sums in computing multiple results. This optimization has multiple effect son improving stencil performance that are particularly important to higher-order stencils: exploits data reuse, reduces floating-point operations, and exposes efficient SIMD parallelism to backend compilers. We study the benefit of this optimization in the context of Geometric Multigrid (GMG), a widely used method to solvePDEs, using four different Jacobi smoothers built from 7-, 13-, 27-and 125-point stencils. We quantify performance, speedup, andnumerical accuracy, and use the Roofline model to qualify our results. Ultimately, we obtain over 4× speedup on the smoothers themselves and up to a 3× speedup on the multigrid solver. Finally, we demonstrate that high-order multigrid solvers have the potential of reducing total data movement and energy by several orders of magnitude.

97 MATHEMATICS AND COMPUTING↗

Machine and Application Aware Partitioning for Adaptive Mesh Refinement Applications. In: HPDC '17 Proceedings of the 26th International Symposium on High-Performance Parallel and Distributed Computing

Load balancing and partitioning are critical when it comes to parallel computations. Popular partitioning strategies based on space filling curves focus on equally dividing work. The partitions produced are independent of the architecture or the application. Given the ever-increasing relative cost of data movement and increasing heterogeneity of our architectures, it is no longer sufficient to only consider an equal partitioning of work. Minimizing communication costs are equally if not more important. Our hypothesis is that an unequal partitioning that minimizes communication costs significantly can scale and perform better than conventional equal-work partitioning schemes. This tradeoff is dependent on the architecture as well as the application. We validate our hypothesis in the context of a finite-element computation utilizing adaptive mesh-refinement. Our central contribution is a new partitioning scheme that minimizes the overall runtime of subsequent computations by performing architecture and application-aware non-uniform work assignment in order to decrease time to solution, primarily by minimizing data-movement. We evaluate our algorithm by comparing it against standard space-filling curve based partitioning algorithms and observing time-to-solution as well as energy-to-solution for solving Finite Element computations on adaptively refined meshes. We demonstrate excellent scalability of our new partition algorithm up to $262,144$ cores on ORNL's Titan and demonstrate that the proposed partitioning scheme reduces overall energy as well as time-to-solution for application codes by up to 22.0%.

Fernando, Milinda↗

Accelerating nuclear configuration interaction calculations through a preconditioned block iterative eigensolver

We describe a number of recently developed techniques for improving the performance of large-scalenuclear configuration interaction calculations on high performance parallel computers. We show thebenefit of using a preconditioned block iterative method to replace the Lanczos algorithm that hastraditionally been used to perform this type of computation. The rapid convergence of the block iterativemethod is achieved by a proper choice of starting guesses of the eigenvectors and the construction of aneffective preconditioner. These acceleration techniques take advantage of special structure of the nuclearconfiguration interaction problem which we discuss in detail. The use of a block method also allows usto improve the concurrency of the computation, and take advantage of the memory hierarchy of modernmicroprocessors to increase the arithmetic intensity of the computation relative to data movement. Wealso discuss the implementation details that are critical to achieving high performance on massivelyparallel multi-core supercomputers, and demonstrate that the new block iterative solver is two to threetimes faster than the Lanczos based algorithm for problems of moderate sizes on a Cray XC30 system.

Shao, Meiyue↗

Work stealing for GPU-accelerated parallel programs in a global address space framework

Task parallelism is an attractive approach to automatically load balance the computation in a parallel system and adapt to dynamism exhibited by parallel systems. Exploiting task parallelism through work stealing has been extensively studied in shared and distributed-memory contexts. In this paper, we study the design of a system that uses work stealing for dynamic load balancing of task-parallel programs executed on hybrid distributed-memory CPU-graphics processing unit (GPU) systems in a global-address space framework. We take into account the unique nature of the accelerator model employed by GPUs, the significant performance difference between GPU and CPU execution as a function of problem size, and the distinct CPU and GPU memory domains. We consider various alternatives in designing a distributed work stealing algorithm for CPU-GPU systems, while taking into account the impact of task distribution and data movement overheads. These strategies are evaluated using microbenchmarks that capture various execution configurations as well as the state-of-the-art CCSD(T) application module from the computational chemistry domain

97 MATHEMATICS AND COMPUTING↗

SharP Data Constructs: Data Constructs to Enable Data-Centric Computing

Extreme-scale applications (i.e., Big-Compute) are becoming increasingly data-intensive, i.e., producing and consuming increasingly large amounts of data. The HPC systems traditionally used for these applications are now used for Big-Data applications such as data analytics, social network analysis, machine learning, and genomics. As a consequence of these trends, the system architecture should be flexible and data-centric. This can already be witnessed in the pre-exascale systems with TBs of on-node hierarchical and heterogeneous memories, PBs of system memory, low-latency, high-throughput networks, and many threaded cores. As such, the pre-exascale systems suit the needs of both Big-Compute and Big-Data applications. Though the system architecture is flexible enough to support both Big-Compute and Big-Data, we argue there is a software gap. Particularly, we need data-centric abstractions to leverage the full potential of the system, i.e., there is a need for native support for data resilience, the ability to express data locality and affinity, mechanisms to reduce data movement, the ability to share data, and abstractions to express User's data usage and data access patterns. In this paper, we (i) show the need for taking a holistic approach towards data-centric abstractions, (ii) show how these approaches were realized in the SHARed data-structure centric Programming abstraction (SharP) library, a data-structure centric programming abstraction, and (iii) apply these approaches to a variety of applications that demonstrate its usefulness. Particularly, we apply these approaches to QMCPack and the Graph500 benchmark and demonstrate the advantages of this approach on extreme-scale systems.

Aderholdt, William Ferrol↗

Productive Programming of Distributed Systems with the SHAD C++ Library

High-performance computing (HPC) is often perceived as a matter of making large-scale systems (e.g., clusters) run as fast as possible, regardless the required programming effort. However, the idea of "bringing HPC to the masses" has recently emerged. Inspired by this vision, we have designed SHAD, the Scalable High-performance Algorithms and Data-structures library. SHAD is open source software, written in C++, for C++ developers. Unlike other HPC libraries for distributed systems, which rely on SPMD models, SHAD adopts a shared-memory programming abstraction, to make C++ programmers feel at home. Underneath, SHAD manages tasking and data-movements, moving the computation where data resides and taking advantage of asynchrony to tolerate network latency. At the bottom of his stack, SHAD can interface with multiple runtime systems: this not only improves developer’s productivity, by hiding the complexity of such software and of the underlying hardware, but also greatly enhance code portability. Thanks to its abstraction layers, SHAD can indeed target different systems, ranging from laptops to HPC clusters, without any need for modifying the user-level code. We have prototyped and open-sourced the implementation of (a subset of) the C++ standard library (STL) targeting multi-node HPC clusters. Our work allows plain STL-based C++ code to scale on HPC systems, with no need for rewriting the code to exploit the complex hardware. SHAD is available under Apache v2 License at https://github.com/pnnl/SHAD. In this paper we overview the design of the SHAD library, depicting its main components: runtime systems abstractions for tasking; parallel and distributed data-structures; STL-compliant interfaces and algorithms.

Castellana, Vito G.↗

Data Federation Challenges in Remote Near-Real-Time Fusion Experiment Data Processing

Fusion energy experiments and simulations provide critical information needed to plan future fusion reactors. As next-generation devices like ITER move toward long-pulse experiments, analyses, including AI and ML, should be performed in a wide range of time and computing constraints, from near-real-time constraints, between-shot analysis, and to campaign-wide long-term analysis. However, the data volume, velocity, and variety make it extremely challenging for analyses using only local computational resources. Researchers need the ability to compose and execute workflows spanning edge resources to large-scale high-performance computing facilities.We present Delta, a system to address data analysis challenges, including AI/ML, in fusion science, by leveraging the ADIOS I/O library and middleware, to support executing science workflows over the wide area network for near-real-time streaming. We discuss the data federation challenges in performing remote workflows, focusing on on-going research work in (1) managing, reducing, and streaming data to minimize I/O and data movement overheads, (2) decompressing and reorganizing data for analysis, and (3) executing workflows for automated data analysis. We introduce examples for deep-learning based data analysis for the fusion domain and demonstrate how we use Delta to construct end-to-end workflows for a fusion device in Korea, connecting a remote DOE facility in the USA. The capability demonstrated by this project is the basis for improving the state of the art for near-real-time data federation amongst remote facilities.

Choi, Jong Youl↗

Performance analysis of distributed symmetric sparse matrix vector multiplication algorithm for multi-core architectures

In this article, sparse matrix vector multiply (SpMVM) is an important kernel that frequently arises in high performance computing applications. Due to its low arithmetic intensity, several approaches have been proposed in literature to improve its scalability and efficiency in large scale computations. In this paper, our target systems are high end multi-core architectures and we use messaging passing interface + open multiprocessing hybrid programming model for parallelism. We analyze the performance of recently proposed implementation of the distributed symmetric SpMVM, originally developed for large sparse symmetric matrices arising in ab initio nuclear structure calculations. We also study important features of this implementation and compare with previously reported implementations that do not exploit underlying symmetry. Our SpMVM implementations leverage the hybrid paradigm to efficiently overlap expensive communications with computations. Our main comparison criterion is the "CPU core hours" metric, which is the main measure of resource usage on supercomputers. We analyze the effects of topology-aware mapping heuristic using simplified network load model. Furthermore, we have tested the different SpMVM implementations on two large clusters with 3D Torus and Dragonfly topology. Our results show that the distributed SpMVM implementation that exploits matrix symmetry and hides communication yields the best value for the "CPU core hours" metric and significantly reduces data movement overheads.

97 MATHEMATICS AND COMPUTING↗

Analytical Modeling of Cache Behavior for Affine Programs

Modern optimizing compiler design integrates program transformation strategies aimed at reducing data movement to/from main memory by exploiting the data cache hierarchy. But in reality, the effect of these transformations on the actual cache miss count is ignored in compilers due to the lack of precise compile-time models of misses for hierarchical caches. The current state of practice for cache miss analysis is based on accurate simulation. However, simulation requires time proportional to the dataset/problem size as well as the number of distinct cache configurations of interest to be evaluated. This paper takes a fundamentally different approach, made possible by focusing on polyhedral programs with static control flow: instead of relying on costly simulation, this paper develops the first closed-form solution for exact modeling of misses in a set-associative cache hierarchy. It enabling simulation-less selection of program transformations at compile-time to optimize cache misses. A push-button tool implementing the approach is developed and used for extensive validation of the complete framework.

affine programs, cache modeling, set associativity↗

Using Xrootd to federate regional storage

While the LHC data movement systems have demonstrated the ability to move data at the necessary throughput, we have identified two weaknesses: the latency for physicists to access data and the complexity of the tools involved. To address these, both ATLAS and CMS have begun to federate regional storage systems using Xrootd. Xrootd, referring to a protocol and implementation, allows us to provide data access to all disk-resident data from a single virtual endpoint. This redirector discovers the actual location of the data and redirects the client to the appropriate site. The approach is particularly advantageous since typically the redirection requires much less than 500 milliseconds and the Xrootd client is conveniently built into LHC physicists analysis tools. Currently, there are three regional storage federations - a US ATLAS region, a European CMS region, and a US CMS region. The US ATLAS and US CMS regions include their respective Tier 1, Tier 2 and some Tier 3 facilities, a large percentage of experimental data is available via the federation. Additionally, US ATLAS has begun studying low-latency regional federations of close-by sites. From the base idea of federating storage behind an endpoint, the implementations and use cases diverge. The CMS software framework is capable of efficiently processing data over high-latency links, so using the remote site directly is comparable to accessing local data. The ATLAS processing model allows a broad spectrum of user applications with varying degrees of performance with regard to latency, a particular focus has been optimizing n-tuple analysis. Both VOs use GSI security. ATLAS has developed a mapping of VOMS roles to specific file system authorizations, while CMS has developed callouts to the site's mapping service. Each federation presents a global namespace to users. For ATLAS, the global-to-local mapping is based on a heuristic-based lookup from the site's local file catalog, while CMS does the mapping based on translations given in a configuration file. We will also cover the latest usage statistics and interesting use cases that have developed over the previous 18 months.

Bauerdick, L.↗

ARCHIE: Data Analysis Acceleration with Array Caching in Hierarchical Storage

© 2018 IEEE. Scientific data analysis typically involves reading massive amounts of data that was generated by simulations, experiments, and observations. Performance of reading such large volumes of data from disk-based file systems is often poor because of the slow and mechanical components in the disks. Recent supercomputing systems are adding non-volatile storage layers in a hierarchy to handle the performance gap between fast main memory and slow disk-based storage. Software libraries for managing this hierarchy not only need efficient reading of data but also reduce user-involvement for cross-layer data movement. Furthermore, these libraries need to support array data access patterns into hierarchical storage management as scientific data is often organized in array-based data structures. Existing software typically manage individual storage layers requiring significant manual process in moving data among them. In this paper, we introduce a new array caching in hierarchical storage (ARCHIE) to accelerate array data analysis in a seamless fashion. ARCHIE evaluates array access patterns and prefetches data with array semantics between storage layers. Our evaluation shows that ARCHIE outperforms state-of-the-art file systems, i.e., Lustre and DataWarp, on a production supercomputing system by up to 5.8× in accessing data by scientific analysis applications.

Dong, Bin↗

Extreme-Scale Distribution-Based Data Analysis

Scientists overview and identify regions of interest by transforming data into compact information descriptors that characterize simulation results and allow detailed analysis on demand. Among many existing feature descriptors, statistical information derived from data samples is a promising approach to taming the big data avalanche because data distributionscomputed from a population can compactly describe the presence and characteristics of salient data features with minimal data movement. The ability to computationally summarize and process data using distributions also provides an efficient and representative capture of information that can adjust to size and resource constraints, with the added benefit that uncertainty associated with results can be quantified and communicated. In this project, we have developed a distribution-based data analysis and visualization framework for optimizing in situ processing and analysis of extreme-scale scientific data.

97 MATHEMATICS AND COMPUTING↗

I/O through High Performance Data Services

The conclusions of this paper are that: (1) moving output data processing to a separate processing area has performance advantages; (2) management of data movement results in large performance gains; (3) the Data Service abstraction enables the development of data staging operations; and (4) using the data service model, we can create adaptive resource aware storage backends.

97 MATHEMATICS AND COMPUTING↗

FTLADS: Object-Logging Based Fault-Tolerant Big Data Transfer System Using Layout Aware Data Scheduling

The layout-aware data scheduling (LADS) data movement framework optimizes congestion for end-to-end data transfers. During data transfer, LADS can avoid congested storage elements by exploiting the underlying storage layout at each endpoint. This improves the I/O bandwidth and hence the data transfer rate across high-speed networks. However, the absence of fault tolerance (FT) in LADS results in data retransmission overhead and may lead to possible data integrity issues upon faults. In this paper, we propose object-logging FT mechanisms to avoid transmitting the objects that are successfully written into the parallel file system (PFS) at the sink end. Depending on the number of log files created for the whole dataset, we have classified our FT mechanisms into three different categories: file logger, transaction logger, and universal logger. Also, to address the space overhead, we have proposed different methods of populating the log files with the information of the successfully transferred objects. We have evaluated the data transfer performance and recovery time overhead of the proposed object-logging-based FT mechanisms on the LADS data transfer framework. Our experimental results reflect that FT mechanisms exhibit negligible overhead (<; 1%) with respect to the data transfer time. Yet, the fault recovery time is 10% higher than the total data transfer time at any fault point.

97 MATHEMATICS AND COMPUTING↗

Accelerating Advanced Light Source Science Through Multi-Facility HPC Workflows

Synchrotron light sources support a wide array of techniques to investigate materials, often producing complex, high-volume data that challenge traditional workflows. At the Advanced Light Source (ALS), we developed infrastructure to move microtomography data over ESnet to ALCF and NERSC, where CPU- and GPU-based algorithms generate 3D reconstructed volumes of experimental samples. We employ two data movement and reconstruction models: real-time processing as data streams directly to NERSC compute nodes, and automated file transfer to NERSC and ALCF file systems. The streaming pipeline provides users with feedback in under ten seconds, while the file-based workflow produces high-quality reconstructions suitable for deeper analysis in 20-30 minutes. This infrastructure enables users to utilize HPC resources without direct access to backend systems. We plan to extend this architecture to more endstations, supporting our beamline scientists and users.

Abramov, David↗

Time-Series Forecast Modeling on High-Bandwidth Network Measurements

With the increasing number of geographically distributed scientific collaborations and the growing sizes of scientific data, it has become challenging for users to achieve the best possible network performance on a shared network. In this paper, we have developed a model to forecast expected bandwidth utilization on high-bandwidth wide area networks. The forecast model can improve the efficiency of the resource utilization and scheduling of data movements on high-bandwidth networks to accommodate ever increasing data volume for large-scale scientific data applications. A univariate time-series forecast model is developed with the Seasonal decomposition of Time series by Loess (STL) and the AutoRegressive Integrated Moving Average (ARIMA) on Simple Network Management Protocol (SNMP) path utilization measurement data. Compared with the traditional approach such as Box-Jenkins methodology to train the ARIMA model, our forecast model reduces computation time up to 92.6 %. It also shows resilience against abrupt network usage changes. Finally, our forecast model conducts the large number of multi-step forecast, and the forecast errors are within the mean absolute deviation (MAD) of the monitored measurements.

97 MATHEMATICS AND COMPUTING↗