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 631 records · Page 35

GSoFa: Scalable Sparse Symbolic LU Factorization on GPUs

Decomposing a matrix $\mathbf {A}$ into a lower matrix $\mathbf {L}$ and an upper matrix $\mathbf {U}$, which is also known as LU decomposition, is an essential operation in numerical linear algebra. For a sparse matrix, LU decomposition often introduces more nonzero entries in the $\mathbf {L}$ and $\mathbf {U}$ factors than in the original matrix. A symbolic factorization step is needed to identify the nonzero structures of $\mathbf {L}$ and $\mathbf {U}$ matrices. Attracted by the enormous potentials of the Graphics Processing Units (GPUs), an array of efforts have surged to deploy various LU factorization steps except for the symbolic factorization, to the best of our knowledge, on GPUs. This article introduces gSoFa, the first GPU-based symbolic factorization design with the following three optimizations to enable scalable LU symbolic factorization for nonsymmetric pattern sparse matrices on GPUs. First, here we introduce a novel fine-grained parallel symbolic factorization algorithm that is well suited for the Single Instruction Multiple Thread (SIMT) architecture of GPUs. Second, we tailor supernode detection into a SIMT friendly process and strive to balance the workload, minimize the communication and saturate the GPU computing resources during supernode detection. Third, we introduce a three-pronged optimization to reduce the excessive space consumption problem faced by multi-source concurrent symbolic factorization. Taken together, gSoFa achieves up to 31× speedup from 1 to 44 Summit nodes (6 to 264 GPUs) and outperforms the state-of-the-art CPU project, on average, by 5×. Notably, gSoFa also achieves up to 47 percent of the peak memory throughput of a V100 GPU in the Summit Supercomputer.

97 MATHEMATICS AND COMPUTING↗

A Programming Model Performance Study Using the NAS Parallel Benchmarks

Harnessing the power of multicore platforms is challenging due to the additional levels of parallelism present. In this paper we use the NAS Parallel Benchmarks to study three programming models, MPI, OpenMP and PGAS to understand their performance and memory usage characteristics on current multicore architectures. To understand these characteristics we use the Integrated Performance Monitoring tool and other ways to measure communication versus computation time, as well as the fraction of the run time spent in OpenMP. The benchmarks are run on two different Cray XT5 systems and an Infiniband cluster. Our results show that in general the three programming models exhibit very similar performance characteristics. In a few cases, OpenMP is significantly faster because it explicitly avoids communication. For these particular cases, we were able to re-write the UPC versions and achieve equal performance to OpenMP. Using OpenMP was also the most advantageous in terms of memory usage. Also we compare performance differences between the two Cray systems, which have quad-core and hex-core processors. We show that at scale the performance is almost always slower on the hex-core system because of increased contention for network resources.

Shan, Hongzhang↗

Livermore Big Artificial Neural Network Toolkit

LBANN is a toolkit that is designed to train artificial neural networks efficiently on high performance computing architectures. It is optimized to take advantages of key High Performance Computing features to accelerate neural network training. Specifically it is optimized for low-latency, high bandwidth interconnects, node-local NVRAM, node-local GPU accelerators, and high bandwidth parallel file systems. It is built on top of the open source Elemental distributed-memory dense and spars-direct linear algebra and optimization library that is released under the BSD license. The algorithms contained within LBANN are drawn from the academic literature and implemented to work within a distributed-memory framework.

Essen, BrianVan↗

Automation of Data Collection Methods for Online Monitoring of Nuclear Power Plants

Maintenance activities in the United States nuclear power industry are mainly centered on preventive and corrective maintenance. Predictive maintenance, which is a variable time basis maintenance strategy that is dependent on the condition of the equipment, is less utilized but potentially more effective. In the current maintenance practices, the nuclear power plants rely on manual processes to supply information to a human decision making process. The plants have several processes that generate information that is not typically used beyond the intended target for collecting that information. This information is a source of data that is manually collected by labor intensive tasks. The data is therefore expensive to collect, yet limited in use. The industry has realized the opportunity present in reducing the labor intensive tasks by automating them and increasing the uses of the data collected by either developing advanced predictive maintenance methods using data-driven decision making that could span from trending to machine learning and artificial intelligence. This shift in maintenance strategy may result in high cost saving impact. In addition to reducing labor requirements and frequency of maintenance activities, this shift in strategy will reduce the materials portion of the operations and maintenance cost as well. The aim of this specific effort is to focus on the automation of data collection process. It is part of a series of planned efforts to target multiple elements of migrating the current preventive maintenance approach in nuclear power plants to a data-driven approach. These elements include data collection, management, analysis, visualization, in addition to value analysis and change enabling. The other five elements of migration were not considered because they are planned to be explicitly studied in multiple parallel efforts. Challenges including regulations requirements or limitations are planned in a future effort that focuses on change enabling. The risk and benefit/cost of the approach shift is planned in another effort targeted towards value. The storage, communication, computational power, data architecture, and protection are part of the information infrastructure and are planned in dedicated data management effort. The best means to present the data for assisted human decision making is planned in a data visualization effort. The data interpretation and patterns recognition for machine decision making is planned through multiple efforts in data analytics over the following years. This work identified seventeen data sources collection methods in nuclear power plants that could be automated for an integrated data platform that enables comprehensive and informed decision making. For each of the identified seventeen data sources, the base, modern, and state of the art states were described. The current state of data collection used by most utilities was identified as the "base state". The current achievable state was identified as the "modern state", a state that has either been achieved by at least some US utilities or in other industries. The future state that requires research and development to reach was identified as the "state of the art" state. The described states used an out-of-the-box thinking approach. This effort found that there is a hug potential for task automation and data fidelity improvements by evolving the current industry data collection methods. The majority of the processes analyzed were found to be manual and labor intensive and fell in the base state. As a result, the modern state often targeted reducing the manual logging efforts by using electronic and semi-intelligent means or complementing the work with sensors and mobile or fixed technologies that can reduce the demand of labor activities. The state of the art state involved applications advanced methods of data correlation, data mining, and machine learning along with increasing the spatial distribution and fidelity of data, coupling multiple data sources, and enabling the use of sophisticated mobile devices and machine comprehension methods.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

A programming model performance study using the NAS parallel benchmarks

Harnessing the power of multicore platforms is challenging due to the additional levels of parallelism present. In this paper we use the NAS Parallel Benchmarks to study three programming models, MPI, OpenMP and PGAS to understand their performance and memory usage characteristics on current multicore architectures. To understand these characteristics we use the Integrated Performance Monitoring tool and other ways to measure communication versus computation time, as well as the fraction of the run time spent in OpenMP. The benchmarks are run on two different Cray XT5 systems and an Infiniband cluster. Our results show that in general the three programming models exhibit very similar performance characteristics. In a few cases, OpenMP is significantly faster because it explicitly avoids communication. For these particular cases, we were able to re-write the UPC versions and achieve equal performance to OpenMP. Using OpenMP was also the most advantageous in terms of memory usage. Also we compare performance differences between the two Cray systems, which have quad-core and hex-core processors.We show that at scale the performance is almost always slower on the hex-core system because of increased contention for network resources.

97 MATHEMATICS AND COMPUTING↗

Livermore Big Artificial Neural Network Toolkit

LBANN is a toolkit that is designed to train artificial neural networks efficiently on high performance computing architectures. It is optimized to take advantages of key High Performance Computing features to accelerate neural network training. Specifically it is optimized for low-latency, high bandwidth interconnects, node-local NVRAM, node-local GPU accelerators, and high bandwidth parallel file systems. It is built on top of the open source Elemental distributed-memory dense and spars-direct linear algebra and optimization library that is released under the BSD license. The algorithms contained within LBANN are drawn from the academic literature and implemented to work within a distributed-memory framework.

Essen, Brian Van↗

PREPARING FOR EXASCALE: ORNL Leadership Computing Application Requirements and Strategy

In 2009 the Oak Ridge Leadership Computing Facility (OLCF), a U.S. Department of Energy (DOE) facility at the Oak Ridge National Laboratory (ORNL) National Center for Computational Sciences (NCCS), elicited petascale computational science requirements from leading computational scientists in the international science community. This effort targeted science teams whose projects received large computer allocation awards on OLCF systems. A clear finding of this process was that in order to reach their science goals over the next several years, multiple projects will require computational resources in excess of an order of magnitude more powerful than those currently available. Additionally, for the longer term, next-generation science will require computing platforms of exascale capability in order to reach DOE science objectives over the next decade. It is generally recognized that achieving exascale in the proposed time frame will require disruptive changes in computer hardware and software. Processor hardware will become necessarily heterogeneous and will include accelerator technologies. Software must undergo the concomitant changes needed to extract the available performance from this heterogeneous hardware. This disruption portends to be substantial, not unlike the change to the message passing paradigm in the computational science community over 20 years ago. Since technological disruptions take time to assimilate, we must aggressively embark on this course of change now, to insure that science applications and their underlying programming models are mature and ready when exascale computing arrives. This includes initiation of application readiness efforts to adapt existing codes to heterogeneous architectures, support of relevant software tools, and procurement of next-generation hardware testbeds for porting and testing codes. The 2009 OLCF requirements process identified numerous actions necessary to meet this challenge: (1) Hardware capabilities must be advanced on multiple fronts, including peak flops, node memory capacity, interconnect latency, interconnect bandwidth, and memory bandwidth. (2) Effective parallel programming interfaces must be developed to exploit the power of emerging hardware. (3) Science application teams must now begin to adapt and reformulate application codes to the new hardware and software, typified by hierarchical and disparate layers of compute, memory and concurrency. (4) Algorithm research must be realigned to exploit this hierarchy. (5) When possible, mathematical libraries must be used to encapsulate the required operations in an efficient and useful way. (6) Software tools must be developed to make the new hardware more usable. (7) Science application software must be improved to cope with the increasing complexity of computing systems. (8) Data management efforts must be readied for the larger quantities of data generated by larger, more accurate science models. Requirements elicitation, analysis, validation, and management comprise a difficult and inexact process, particularly in periods of technological change. Nonetheless, the OLCF requirements modeling process is becoming increasingly quantitative and actionable, as the process becomes more developed and mature, and the process this year has identified clear and concrete steps to be taken. This report discloses (1) the fundamental science case driving the need for the next generation of computer hardware, (2) application usage trends that illustrate the science need, (3) application performance characteristics that drive the need for increased hardware capabilities, (4) resource and process requirements that make the development and deployment of science applications on next-generation hardware successful, and (5) summary recommendations for the required next steps within the computer and computational science communities.

applications requirements↗

Record acceleration of the two-dimensional Ising model using a high-performance wafer-scale engine

The versatility and wide-ranging applicability of the Ising model, originally introduced to study phase transitions in magnetic materials, have made it a cornerstone in statistical physics and a valuable tool for evaluating the performance of emerging computer hardware. Here, we present a novel implementation of the two-dimensional Ising model on Cerebras Wafer-Scale Engine (WSE) – a revolutionary processor that is opening new frontiers in computing. In our deployment of the checkerboard algorithm, we optimized the Ising model to take advantage of the unique WSE architecture. Specifically, we employed a compressed bit representation storing 16 spins on each int16 word, and efficiently distributed the spins over the processing units enabling seamless weak scaling and limiting communications to only immediate neighboring units. Our implementation can handle up to 754 simulations in parallel, achieving an aggregate of over 61.8 trillion flip attempts per second for Ising models with up to 200 million spins. This represents a gain of up to 148 times over previously reported single-devices with a highly optimized implementation on NVIDIA V100 and up to 88 times in productivity compared to NVIDIA H100. Our findings highlight the significant potential of the WSE in scientific computing, particularly in the field of materials modeling.

Ising model↗

Super instruction architecture of petascale electronic structure software: the story

Theoretical methods in chemistry lead to algorithms for the computation of electronic energies and other properties of electronic wave functions that require large numbers of floating point operations and involve large data sets. Thus, computational chemists are very interested in using massively parallel computer systems and in particular the new petascale systems. In this paper we discuss a new programming paradigm that was developed at the Quantum Theory Project to construct electronic structure software that can scale to large numbers of cores of the order of 100,000 and beyond to solve problems in materials engineering relevant to the problems facing society today.

Chemistry↗

Toward a persistent event-streaming system for high-performance computing applications

High-performance computing (HPC) applications have traditionally relied on parallel file systems and file transfer services to manage data movement and storage. Alternative approaches have been proposed that use direct communications between application components, trading persistence and fault tolerance for speed. Event-driven architectures, as popularized in enterprise contexts, present a compelling middle ground, avoiding the performance cost and API constraints of parallel file systems while retaining persistence and offering impedance matching between application components. However, adapting streaming frameworks to HPC workloads requires addressing challenges unique to HPC systems. This paper investigates the potential for a streaming framework designed for HPC infrastructures and use cases. We introduce Mofka, a persistent event-streaming framework designed specifically for HPC environments. Mofka combines the capabilities of a traditional streaming service with optimizations tailored to the HPC context, such as support for massively multicore nodes, efficient scaling for large producer-consumer workflows, RDMA-enabled high-performance network communications, specialized network fabrics with multiple links per node, and efficient handling of large scientific data payloads. Built using the Mochi suite of HPC data service components, Mofka provides a lightweight, modular, and high-performance solution for persistent streaming in HPC systems. We present the architecture of Mofka and evaluate its performance against Kafka and Redpanda using benchmarks on diverse platforms, including Argonne's Polaris and Oak Ridge's Frontier supercomputers, showing up to 8× improvement in throughput in some scenarios. We then demonstrate its utility in several real-world applications: a tomographic reconstruction pipeline, a workflow for the discovery of metal-organic frameworks for carbon capture, and the instrumentation of Dask workflows for provenance tracking and performance analysis.

HPC↗

A Runtime Environment for Supporting Research in Resilient HPC System Software & Tools

The high-performance computing (HPC) community continues to increase the size and complexity of hardware platforms that support advanced scientific workloads. The runtime environment (RTE) is a crucial layer in the software stack for these large-scale systems. The RTE manages the interface between the operating system and the application running in parallel on the machine. The deployment of applications and tools on large-scale HPC computing systems requires the RTE to manage process creation in a scalable manner, support sparse connectivity, and provide fault tolerance. We have developed a new RTE that provides a basis for building distributed execution environments and developing tools for HPC to aid research in system software and resilience. This paper describes the software architecture of the Scalable runTime Component Infrastructure (STCI), which is intended to provide a complete infrastructure for scalable start-up and management of many processes in large-scale HPC systems. We highlight features of the current implementation, which is provided as a system library that allows developers to easily use and integrate STCI in their tools and/or applications. The motivation for this work has been to support ongoing research activities in fault-tolerance for large-scale systems. We discuss the advantages of the modular framework employed and describe two use cases that demonstrate its capabilities: (i) an alternate runtime for a Message Passing Interface (MPI) stack, and (ii) a distributed control and communication substrate for a fault-injection tool.

Computer Science↗

Porting DMRG++ Scientific Application to OpenPOWER

With the rapidly changing microprocessor designs and architectural diversity (multi-cores, many-cores, accelerators) for the next generation HPC systems, scientific applications must adapt to the hardware, to exploit the different types of parallelism and resources available in the architecture. To get the benefit of all the in-node hardware threads, it is important to use a single programming model to map and coordinate the available work to the different heterogeneous execution units in the node (e.g., multi-core hardware threads (latency optimized), accelerators (bandwidth optimized), etc.).Our goal is to show that we can manage the node complexity of these systems by using OpenMP for in-node parallelization by exploiting different “programming styles” supported by OpenMP 4.5 to program CPU cores and accelerators. Finding out the suitable programming-style (e.g., SPMD style, multi-level tasks, accelerator programming, nested parallelism, or a combination of these) using the latest features of OpenMP to maximize performance and achieve performance portability across heterogeneous and homogeneous systems is still an open research problem.We developed a mini-application, Kronecker Product (KP), from the original DMRG++ application (sparse matrix algebra) computational motif to experiment with different OpenMP programming styles on an OpenPOWER architecture and present their results in this paper.

Chatterjee, Ronnie↗

ASC-ATDM Performance Portability Requirements for 2015-2019

This report outlines the research, development, and support requirements for the Advanced Simulation and Computing (ASC ) Advanced Technology, Development, and Mitigation (ATDM) Performance Portability (a.k.a., Kokkos) project for 2015 - 2019 . The research and development (R&D) goal for Kokkos (v2) has been to create and demonstrate a thread - parallel programming model a nd standard C++ library - based implementation that enables performance portability across diverse manycore architectures such as multicore CPU, Intel Xeon Phi, and NVIDIA Kepler GPU. This R&D goal has been achieved for algorithms that use data parallel pat terns including parallel - for, parallel - reduce, and parallel - scan. Current R&D is focusing on hierarchical parallel patterns such as a directed acyclic graph (DAG) of asynchronous tasks where each task contain s nested data parallel algorithms. This five y ear plan includes R&D required to f ully and performance portably exploit thread parallelism across current and anticipated next generation platforms (NGP). The Kokkos library is being evaluated by many projects exploring algorithm s and code design for NGP. Some production libraries and applications such as Trilinos and LAMMPS have already committed to Kokkos as their foundation for manycore parallelism an d performance portability. These five year requirements includes support required for current and antic ipated ASC projects to be effective and productive in their use of Kokkos on NGP. The greatest risk to the success of Kokkos and ASC projects relying upon Kokkos is a lack of staffing resources to support Kokkos to the degree needed by these ASC projects. This support includes up - to - date tutorials, documentation, multi - platform (hardware and software stack) testing, minor feature enhancements, thread - scalable algorithm consulting, and managing collaborative R&D.

97 MATHEMATICS AND COMPUTING↗

MPI + OpenACC: Accelerating radiation transport mini-application, minisweep, on heterogeneous systems

According to the article, architectures are rapidly evolving, and exascale machines are expected to offer billion-way concurrency. We need to rethink algorithms, languages and programming models among other components in order to migrate large scale applications and explore parallelism on these machines. Although directive-based programming models allow programmers to worry less about programming and more about science, expressing complex parallel patterns in these models can be a daunting task especially when the goal is to match the performance that the hardware platforms can offer. One such pattern is wavefront. This paper extensively studies a wavefront-based miniapplication for Denovo, a production code for nuclear reactor modeling. We parallelize the Koch–Baker–Alcouffe (KBA) parallel-wavefront sweep algorithm in the main kernel of Minisweep (the miniapplication) using CUDA 9.0, OpenMP 4.0 (SIMD) and OpenACC 2.6. Our OpenACC implementation running on NVIDIA’s next-generation Volta GPU boasts an 85.06x speedup over serial code, which is larger than CUDA’s 83.72x speedup over the same serial implementation. We also explore the scalability of our solution using MPI to decompose our simulation domain, allowing us to run on many nodes and accelerators present in state-of-the-art HPC systems. Our parallelization effort across platforms also motivated us to define an abstract parallelism model that is architecture independent, with a goal of creating software abstractions that can be used by applications employing the wavefront sweep motif.

Computer Science↗

Parallelizable adjoint stencil computations using transposed forward-mode algorithmic differentiation

Algorithmic differentiation (AD) is a tool for generating discrete adjoint solvers, which efficiently compute gradients of functions with many inputs, for example for use in gradient-based optimization. AD is often applied to large computations such as stencil operators, which are an important part of most structured-mesh PDE solvers. Stencil computations are often parallelized, for example by using OpenMP, and optimized by using techniques such as cache-blocking and tiling to fully utilize multicore CPUs and many-core accelerators and GPUs. Differentiating these codes with conventional reverse-mode AD results in adjoint codes that cannot be expressed as stencil operations and may not be easily parallelizable. They thus leave most of the compute power of modern architectures unused. We present a method that combines forward-mode AD and loop transformation to generate adjoint solvers that use the same memory access pattern as the original computation that they are derived from and can benefit from the same optimization techniques. The effectiveness of this method is demonstrated by generating a scalable adjoint CFD solver for multicore CPUs and Xeon Phi accelerators.

algorithmic differentiation↗

Real-Time Stochastic Optimization of Complex Energy Systems on High-Performance Computers

A scalable approach computes in operationally-compatible time the energy dispatch under uncertainty for electrical power grid systems of realistic size with thousands of scenarios. The authors propose several algorithmic and implementation advances in their parallel solver PIPS-IPM for stochastic optimization problems. New developments include a novel, incomplete, augmented, multicore, sparse factorization implemented within the PARDISO linear solver and new multicore- and GPU-based dense matrix implementations. They show improvement on the interprocess communication on Cray XK7 and XC30 systems. PIPS-IPM is used to solve 24-hour horizon power grid problems with up to 1.95 billion decision variables and 1.94 billion constraints on Cray XK7 and Cray XC30, with observed parallel efficiencies and solution times within an operationally defined time interval. To the authors' knowledge, "real-time"-compatible performance on a broad range of architectures for this class of problems hasn't been possible prior to this work

Computer Science↗

Active Flash: Performance-Energy Tradeoffs for Out-of-Core Processing on Non-Volatile Memory Devices

In this abstract, we study the performance and energy tradeoffs involved in migrating data analysis into the flash device, a process we refer to as Active Flash. The Active Flash paradigm is similar to 'active disks', which has received considerable attention. Active Flash allows us to move processing closer to data, thereby minimizing data movement costs and reducing power consumption. It enables true out-of-core computation. The conventional definition of out-of-core solvers refers to an approach to process data that is too large to fit in the main memory and, consequently, requires access to disk. However, in Active Flash, processing outside the host CPU literally frees the core and achieves real 'out-of-core' analysis. Moving analysis to data has long been desirable, not just at this level, but at all levels of the system hierarchy. However, this requires a detailed study on the tradeoffs involved in achieving analysis turnaround under an acceptable energy envelope. To this end, we first need to evaluate if there is enough computing power on the flash device to warrant such an exploration. Flash processors require decent computing power to run the internal logic pertaining to the Flash Translation Layer (FTL), which is responsible for operations such as address translation, garbage collection (GC) and wear-leveling. Modern SSDs are composed of multiple packages and several flash chips within a package. The packages are connected using multiple I/O channels to offer high I/O bandwidth. SSD computing power is also expected to be high enough to exploit such inherent internal parallelism within the drive to increase the bandwidth and to handle fast I/O requests. More recently, SSD devices are being equipped with powerful processing units and are even embedded with multicore CPUs (e.g. ARM Cortex-A9 embedded processor is advertised to reach 2GHz frequency and deliver 5000 DMIPS; OCZ RevoDrive X2 SSD has 4 SandForce controllers, each with 780MHz max frequency Tensilica core). Efforts that take advantage of the available computing cycles on the processors on SSDs to run auxiliary tasks other than actual I/O requests are beginning to emerge. Kim et al. investigate database scan operations in the context of processing on the SSDs, and propose dedicated hardware logic to speed up scans. Also, cluster architectures have been explored, which consist of low-power embedded CPUs coupled with small local flash to achieve fast, parallel access to data. Processor utilization on SSD is highly dependent on workloads and, therefore, they can be idle during periods with no I/O accesses. We propose to use the available processing capability on the SSD to run tasks that can be offloaded from the host. This paper makes the following contributions: (1) We have investigated Active Flash and its potential to optimize the total energy cost, including power consumption on the host and the flash device; (2) We have developed analytical models to analyze the performance-energy tradeoffs for Active Flash, by treating the SSD as a blackbox, this is particularly valuable due to the proprietary nature of the SSD internal hardware; and (3) We have enhanced a well-known SSD simulator (from MSR) to implement 'on-the-fly' data compression using Active Flash. Our results provide a window into striking a balance between energy consumption and application performance.

42 ENGINEERING↗

Symplectic multi-particle tracking on GPUs

A symplectic multi-particle tracking model is implemented on the Graphic Processing Units (GPUs) using the Compute Unified Device Architecture (CUDA) language. The symplectic tracking model can preserve phase space structure and reduce non-physical effects in long term simulation, which is important for beam property evaluation in particle accelerators. Though this model is computationally expensive, it is very suitable for parallelization and can be accelerated significantly by using GPUs. In this paper, we optimized the implementation of the symplectic tracking model on both single GPU and multiple GPUs. Using a single GPU processor, the code achieves a factor of 2-10 speedup for a range of problem sizes compared with the time on a single state-of-the-art Central Processing Unit (CPU) node with similar power consumption and semiconductor technology. It also shows good scalability on a multi-GPU cluster at Oak Ridge Leadership Computing Facility. In an application to beam dynamics simulation, the GPU implementation helps save more than a factor of two total computing time in comparison to the CPU implementation.

43 PARTICLE ACCELERATORS↗