Search NASA⌕ Search

SEARCH · Search NASA

Results for “Runtime”

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 379 records · Page 21

MatRIS: Addressing the Challenges for Portability and Heterogeneity Using Tasking for Matrix Decomposition (Cholesky)

The ubiquitous in-node heterogeneity of HPC and cloud computing platforms makes software portability and performance optimization extremely challenging. Described here, the MatRIS multilevel math library abstraction framework employs tasking to alleviate these difficulties. MatRIS includes the IRIS task-based runtime on the bottom level and exposes different layers of abstraction to render algorithms architecturally agnostic. MatRIS ensures the decomposition and creation of tasks that represent the necessary encapsulation of the optimized kernels from both vendor and open-source math libraries. Once built, MatRIS can select different combinations of accelerators at runtime, making it portable even on diverse heterogeneous architectures. By leveraging the IRIS runtime’s features for managing heterogeneity, MatRIS deploys algorithms that remove the need to specify orchestration and data transfer. This study describes how the serial task abstraction of a tiled Cholesky factorization is made portable and scalable in the case of multi-device and multi-vendor heterogeneity on a node with NVIDIA and AMD GPUs by using MatRIS. First, we demonstrate that Cholesky in MatRIS provides multi-GPU scalability that offers competitive performance versus cuSolverMG. Then, we present the challenges and opportunities for heterogeneous execution.

Monil, M. A. H.↗

Quality Control of Additively Manufactured Metallic Structures with Machine Learning of Thermography Images

Additive manufacturing (AM) of high-strength metals is currently based on the laser powder bed fusion (LPBF) process, which can introduce internal material flaws, such as pores and anisotropy. Quality control (QC) requires nondestructive evaluation of actual AM structures. Flash thermography is a potentially promising QC technique because it is scalable to arbitrary structure size. However, the detection sensitivity of this method is limited by noise. We investigate separation of signal from noise in thermography images using several machine learning (ML) methods, including new spatial–temporal blind source separation and spatial–temporal sparse dictionary learning methods. Performance of the ML methods is benchmarked using thermography data obtained from imaging stainless steel 316L and Inconel 718 specimens produced by the LPBF method with imprinted calibrated porosity defects. The ML methods are ranked by F-score and execution runtime. Finally, the ML methods with higher accuracy require a longer runtime. However, this runtime is sufficiently short to perform QC within a realistic time frame.

36 MATERIALS SCIENCE↗

Spatial decomposition of structured grids for nuclear reactor simulations

Spatial decomposition methods based on graph partitioning are developed and implemented in the high fidelity neutron transport code MPACT. These graph-based spatial decomposition methods are more general than previous decomposition methods and typically provide better load balance and reduced runtimes due to their improved parallel efficiency. Correlations are drawn between simulation runtime and the balance of the partition for 2D simulations. Comparisons are made using partition balance metrics for different decomposition schemes in 2D and 3D simulations. Finally, for typical ranges of subdomains, graph-based partitioning methods offer significant reductions to runtimes. However, for highly decomposed problems, these graph-based methods may decrease convergence rates, thus reducing parallel efficiency compared to older methods.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Impact of refrigerant undercharge faults on building indoor conditions and HVAC system operation in residential Buildings: A simulation study

This study investigates the impact of refrigerant undercharge on indoor temperature and HVAC system performance in residential buildings. Simulation models for typical residential buildings in Orlando, FL and Indianapolis, IN were developed using the ResStock database. A refrigerant undercharge fault model was then applied to the simulations with varying levels of fault intensity. The paper offers an extensive analysis, revealing that variations in supply air temperature, equipment runtime, and cooling energy consumption due to the level of refrigerant undercharge faults are notably significant on a summer representative day. Similarly, on a winter representative day, changes in supply air temperature and runtime are significant as well as changes in supplemental heat energy consumption. We find that occupants may remain oblivious to these faults during the cooling season, particularly when the HVAC system is oversized; in that case, supply air temperature data could help detect a fault. Another challenge is that during the heating season, when the supplemental heater operates, it is difficult to identify a refrigerant undercharge fault using only indoor and supply air temperature data. Finally, this study finds that supply air temperature, equipment runtime, and supplemental heater energy consumption data can help in detecting refrigerant undercharge faults.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

On the adequacy of lightweight thread approaches for high-level parallel programming models

High-level parallel programming models (PMs) are becoming crucial in order to extract the computational power of current on-node multi-threaded parallelism. The most popular PMs, such as OpenMP or OmpSs, are directive-based: the complexity of the hardware is hidden by the underlying runtime system, improving coding productivity. The implementations of OpenMP usually rely on POSIX threads (pthreads), offering excellent performance for coarse-grained parallelism and a perfect match with the current hardware. OmpSs is a task oriented PM based on an ad hoc runtime solution called Nanos++; it is the precursor of the tasking parallelism in the OpenMP tasking specification. A recent trend in runtimes and applications points to leveraging massive on-node parallelism in conjunction with fine-grained and dynamic scheduling paradigms. In this paper we analyze the behavior of the OpenMP and OmpSs PMs on top of the recently emerged Generic Lightweight Threads (GLT) API. GLT exposses a common API for lightweight thread (LWT) libraries that offers the possibility of running the same application over different native LWT solutions. We describe the design details of those high-level PMs implemented on top of GLT and analyze different scenarios in order to assess where the use of LWTs may benefit application performance. Our work reveals those scenarios where LWTs overperform pthread-based solutions and compares the performance between an ad hoc solution and a generic implementation.

97 MATHEMATICS AND COMPUTING↗

An exploration of online-simulation-driven portfolio scheduling in Workflow Management Systems

Workflow Management Systems used to automate the execution of scientific workflow applications on parallel and distributed computing platforms must make scheduling decisions at runtime. A large number of workflow scheduling algorithms have been proposed in the literature, but often these algorithms are evaluated based on simplifying assumptions that may not hold in practice. Furthermore, published algorithm evaluation and/or comparison results are necessarily only for a subset of all possible scenarios, and thus may not include scenarios relevant to particular use-cases. Consequently, it is difficult for Workflow Management Systems (WMSs) developers to decide which scheduling algorithm should be implemented. To obviate this difficulty, one possible approach is to implement a portfolio of scheduling algorithms and select the most effective algorithm at runtime. One method for performing this selection is to run an online simulation for each algorithm in the portfolio. The algorithm that leads to the best performance, in simulation, is selected for future use. The above simulation-driven portfolio scheduling (SDPS) approach has been proposed in a few parallel and distributed computing contexts. The main objective of this work is to evaluate the feasibility and potential merit of SDPS if implemented in WMSs. Here we perform this evaluation using simulated WMS executions, where the simulations are instantiated from real-world platform and workflow configurations. Our main finding is that SDPS is on par with or outperforms an approach in which a single algorithm is used, where this algorithm is the one that performs best on average across all our experimental scenarios. Furthermore, we find that SDPS remains an attractive proposition even in the presence of high levels of simulation error and for simulators with relatively low levels of sophistication. In many of our experimental scenarios we find that mitigating simulation error at runtime can further improve performance. Finally, we show that simulation overhead can be made sufficiently low for SDPS to be feasible in practice.

97 MATHEMATICS AND COMPUTING↗

The Metropolis Monte Carlo method with CUDA enabled Graphic Processing Units

We present a CPU–GPU system for runtime acceleration of large molecular simulations using GPU computation and memory swaps. The memory architecture of the GPU can be used both as container for simulation data stored on the graphics card and as floating-point code target, providing an effective means for the manipulation of atomistic or molecular data on the GPU. To fully take advantage of this mechanism, efficient GPU realizations of algorithms used to perform atomistic and molecular simulations are essential. Our system implements a versatile molecular engine, including inter-molecule interactions and orientational variables for performing the Metropolis Monte Carlo (MMC) algorithm, which is one type of Markov chain Monte Carlo. By combining memory objects with floating-point code fragments we have implemented an MMC parallel engine that entirely avoids the communication time of molecular data at runtime. Our runtime acceleration system is a forerunner of a new class of CPU–GPU algorithms exploiting memory concepts combined with threading for avoiding bus bandwidth and communication. The testbed molecular system used here is a condensed phase system of oligopyrrole chains. A benchmark shows a size scaling speedup of 60 for systems with 210,000 pyrrole monomers. Our implementation can easily be combined with MPI to connect in parallel several CPU–GPU duets. -- Highlights: •We parallelize the Metropolis Monte Carlo (MMC) algorithm on one CPU—GPU duet. •The Adaptive Tempering Monte Carlo employs MMC and profits from this CPU—GPU implementation. •Our benchmark shows a size scaling-up speedup of 62 for systems with 225,000 particles. •The testbed involves a polymeric system of oligopyrroles in the condensed phase. •The CPU—GPU parallelization includes dipole—dipole and Mie—Jones classic potentials.

97 MATHEMATICS AND COMPUTING↗

Clustering huge protein sequence sets in linear time

Metagenomic datasets contain billions of protein sequences that could greatly enhance largescale functional annotation and structure prediction. Utilizing this enormous resource would require reducing its redundancy by similarity clustering. However, clustering hundreds of millions of sequences is impractical using current algorithms because their runtimes scale as the input set size N times the number of clusters K, which is typically of similar order as N, resulting in runtimes that increase almost quadratically with N. We developed Linclust, the first clustering algorithm whose runtime scales as N, independent of K. It can also cluster datasets several times larger than the available main memory. We cluster 1.6 billion metagenomic sequence fragments in 10 h on a single server to 50% sequence identity, >1000 times faster than has been possible before. Linclust will help to unlock the great wealth contained in metagenomic and genomic sequence databases.

59 BASIC BIOLOGICAL SCIENCES↗

Towards Low-Overhead Resilience for Data Parallel Deep Learning

Data parallel techniques have been widely adopted both in academia and industry as a tool to enable scalable training of deep learning models. At scale, DL training jobs can fail due to software or hardware bugs, may need to be preempted or terminated due to unexpected events, or may perform suboptimally because they were misconfigured. Under such circumstances, there is a need to recover and/or reconfigure data-parallel DL training jobs on-the-fly, while minimizing the impact on the accuracy of the DNN model and the runtime overhead. In this regard, state-of-art techniques adopted by the HPC community mostly rely on checkpoint-restart, which inevitably leads to loss of progress, thus increasing the runtime overhead. In this paper we explore alternative techniques that exploit the properties of modern deep learning frameworks (overlapping of gradient averaging and weight updates with local gradient computations through pipeline parallelism) to reduce the overhead of resilience/elasticity. To this end we introduce a failure simulation framework and two resilience strategies (immediate mini-batch rollback and lossy forward recovery), which we study compared with checkpoint-restart approaches in a variety of settings in order to understand the trade-offs between the accuracy loss of the DNN model and the runtime overhead.

data-parallel training↗

An Approach for Detecting Inconsistencies between Behavioral Models of the Software Architecture and the Code

In practice, inconsistencies between architectural documentation and the code might arise due to improper implementation of the architecture or the separate, uncontrolled evolution of the code. Several approaches have been proposed to detect the inconsistencies between the architecture and the code but these tend to be limited for capturing inconsistencies that might occur at runtime. We present a runtime verification approach for detecting inconsistencies between the dynamic behavior of the architecture and the actual code. The approach is supported by a set of tools that implement the architecture and the code patterns in Prolog, and support the automatic generation of runtime monitors for detecting inconsistencies. We illustrate the approach and the toolset for a Crisis Management System case study.

Code generation↗

A Task Based Approach for Co-Scheduling Ensemble Workloads on Heterogeneous Nodes

Scientific workflows consist of multiple, connected applications, with data and results flowing from one to another in a pipeline. Traditionally, such workflows are executed in sequential order, storing intermediate data in storage disks. Co-scheduling application workflows concurrently on the same compute nodes would greatly reduce the cost of moving data to/from storage and allow real-time analysis of intermediate results. Nevertheless, most parallel programming runtimes do not allow seamless integration of various applications in a scientific workflow, in part due to the complexity of managing data and resources. The situation is even more complicated for heterogeneous systems. In this work we extend the Minos Computing Library (MCL) runtime to accelerate pipe-lined and parallel workloads where multiple applications are running in the same system. MCL’s asynchronous task library and runtime dynamically manages resources to allow co-scheduling of multiple processes sharing heterogeneous resources. In addition, we design a custom ex- tension of the Open Compute Language (OpenCL) to enable multiple processes to share device memory. We enable MCL to coordinate these shared buffers to allow for easy, fast data sharing between applications. Using malleable micro-benchmarks and two application workflows that combine scientific simulation and AI-based analysis, we show that our method outperforms traditional approaches.

Index Terms—Parallel systems, Scheduling and Task ↗

A domain-specific compiler for a parallel multiresolution adaptive numerical simulation environment

This paper describes the design and implementation of a layered domain-specific compiler to support MADNESS---Multiresolution ADaptive Numerical Environment for Scientific Simulation. MADNESS is a high-level software environment for the solution of integral and differential equations in many dimensions, using adaptive and fast harmonic analysis methods with guaranteed precision. MADNESS uses k-d trees to represent spatial functions and implements operators like addition, multiplication, differentiation, and integration on the numerical representation of functions. The MADNESS runtime system provides global namespace support and a task-based execution model including futures. MADNESS is currently deployed on massively parallel supercomputers and has enabled many science advances. Due to the highly irregular and statically unpredictable structure of the k-d trees representing the spatial functions encountered in MADNESS applications, only purely runtime approaches to optimization have previously been implemented in the MADNESS framework. This paper describes a layered domain-specific compiler developed to address some performance bottlenecks in MADNESS. The newly developed static compile-time optimizations, in conjunction with the MADNESS runtime support, enable significant performance improvement for the MADNESS framework.

Rajbhandari, Samyam↗

A performance model for GPUs with caches

To exploit the abundant computational power of the world's fastest supercomputers, an even workload distribution to the typically heterogeneous compute devices is necessary. While relatively accurate performance models exist for conventional CPUs, accurate performance estimation models for modern GPUs do not exist. This paper presents two accurate models for modern GPUs: a sampling-based linear model, and a model based on machine-learning (ML) techniques which improves the accuracy of the linear model and is applicable to modern GPUs with and without caches. We first construct the sampling-based linear model to predict the runtime of an arbitrary OpenCL kernel. Based on an analysis of NVIDIA GPUs' scheduling policies we determine the earliest sampling points that allow an accurate estimation. The linear model cannot capture well the significant effects that memory coalescing or caching as implemented in modern GPUs have on performance. We therefore propose a model based on ML techniques that takes several compiler-generated statistics about the kernel as well as the GPU's hardware performance counters as additional inputs to obtain a more accurate runtime performance estimation for modern GPUs. We demonstrate the effectiveness and broad applicability of the model by applying it to three different NVIDIA GPU architectures and one AMD GPU architecture. On an extensive set of OpenCL benchmarks, on average, the proposed model estimates the runtime performance with less than 7 percent error for a second-generation GTX 280 with no on-chip caches and less than 5 percent for the Fermi-based GTX 580 with hardware caches. On the Kepler-based GTX 680, the linear model has an error of less than 10 percent. On an AMD GPU architecture, Radeon HD 6970, the model estimates with 8 percent of error rates. As a result, the proposed technique outperforms existing models by a factor of 5 to 6 in terms of accuracy.

97 MATHEMATICS AND COMPUTING↗

Algorithm-Based Fault Tolerance for Convolutional Neural Networks

Convolutional neural networks (CNNs) are becoming more and more important for solving challenging and critical problems in many fields. CNN inference applications have been deployed in safety-critical systems, which may suffer from soft errors caused by high-energy particles, high temperature, or abnormal voltage. Of critical importance is ensuring the stability of the CNN inference process against soft errors. Traditional fault tolerance methods are not suitable for CNN inference because error-correcting code is unable to protect computational components, instruction duplication techniques incur high overhead, and existing algorithm-based fault tolerance (ABFT) techniques cannot protect all convolution implementations. In this paper, we focus on how to protect the CNN inference process against soft errors as efficiently as possible, with the following three contributions. (1) We propose several systematic ABFT schemes based on checksum techniques and analyze their fault protection ability and runtime thoroughly. Unlike traditional ABFT based on matrix-matrix multiplication, our schemes support any convolution implementations. (2) We design a novel workflow integrating all the proposed schemes to obtain a high detection/correction ability with limited total runtime overhead. (3) We perform our evaluation using ImageNet with well-known CNN models including AlexNet, VGG-19, ResNet-18, and YOLOv2. Here, experimental results demonstrate that our implementation can handle soft errors with very limited runtime overhead (4%~8% in both error-free and error-injected situations).

97 MATHEMATICS AND COMPUTING↗

Barrier elision for production parallel programs

Large scientific code bases are often composed of several layers of runtime libraries, implemented in multiple programming languages. In such situation, programmers often choose conservative synchronization patterns leading to suboptimal performance. In this paper, we present context-sensitive dynamic optimizations that elide barriers redundant during the program execution. In our technique, we perform data race detection alongside the program to identify redundant barriers in their calling contexts; after an initial learning, we start eliding all future instances of barriers occurring in the same calling context. We present an automatic on-the-fly optimization and a multi-pass guided optimization. We apply our techniques to NWChem - a 6 million line computational chemistry code written in C/C++/Fortran that uses several runtime libraries such as Global Arrays, ComEx, DMAPP, and MPI. Our technique elides a surprisingly high fraction of barriers (as many as 63%) in production runs. This redundancy elimination translates to application speedups as high as 14% on 2048 cores. Our techniques also provided valuable insight about the application behavior, later used by NWChem developers. Overall, we demonstrate the value of holistic context-sensitive analyses that consider the domain science in conjunction with the associated runtime software stack.

97 MATHEMATICS AND COMPUTING↗

Efficient Parallel Discrete Event Simulation on Cloud/Virtual Machine Platforms

Cloud and Virtual machine (VM) technologies present new challenges with respect to performance and monetary cost in executing parallel discrete event simulation (PDES) applications. Due to the introduction of overall cost as a metric, the traditional use of the highest-end computing configuration is no longer the most obvious choice. Moreover, the unique runtime dynamics and configuration choices of Cloud and VM platforms introduce new design considerations and runtime characteristics specific to PDES over Cloud/VMs. Here, an empirical study is presented to guide an understanding of the dynamics, trends, and trade-offs in executing PDES on Cloud/VM platforms. Performance and cost measures obtained from multiple PDES applications executed on the Amazon EC2 Cloud and on a high-end VM host machine reveal new, counterintuitive VM--PDES dynamics and guidelines. One of the critical aspects uncovered is the fundamental mismatch in hypervisor scheduler policies designed for general cloud workloads versus the virtual time ordering needed for PDES workloads. This insight is supported by experimental data revealing the gross deterioration in PDES performance traceable to VM scheduling policy. To overcome this fundamental problem, the design and implementation of a new deadlock-free scheduler algorithm are presented, optimized specifically for PDES applications on VMs. The scalability of our scheduler has been tested up to 128 VMs multiplexed on 32 cores, showing significant improvement in the runtime relative to the default Cloud/VM scheduler. The observations, algorithmic design, and results are timely for emerging cloud/VM-based installations, highlighting the need for PDES-specific support in high performance discrete event simulations on Cloud/VM platforms.

97 MATHEMATICS AND COMPUTING↗

Static and Dynamic Frequency Scaling on Multicore CPUs

Dynamic voltage and frequency scaling (DVFS) adapts CPU power consumption by modifying a processor’s operating frequency (and the associated voltage). Typical approaches employing DVFS involve default strategies such as running at the lowest or the highest frequency, or observing the CPU’s runtime behavior and dynamically adapting the voltage/frequency configuration based on CPU usage. In this paper, we argue that many previous approaches suffer from inherent limitations, such as not account- ing for processor-specific impact of frequency changes on energy for different workload types. We first propose a lightweight runtime-based approach to automatically adapt the frequency based on the CPU workload, that is agnostic of the processor characteristics. We then show that further improvements can be achieved for affine kernels in the application, using a compile-time characterization instead of run-time monitoring to select the frequency and number of CPU cores to use. Our framework relies on a one-time energy characterization of CPU-specific DVFS profiles followed by a compile-time categorization of loop-based code segments in the application. These are combined to determine a priori of the frequency and the number of cores to use to execute the application so as to optimize energy or energy-delay product, outperforming runtime approach. Extensive evaluation on 60 benchmarks and five multi-core CPUs show that our approach systematically outperforms the powersave Linux governor, while improving overall performance.

24 POWER TRANSMISSION AND DISTRIBUTION↗

An LLVM Instrumentation Plug-in for Score-P

Reducing application runtime, scaling parallel applications to higher numbers of processes/threads, and porting applications to new hardware architectures are tasks necessary in the software development process. Therefore, developers have to investigate and understand application runtime behavior. Tools such as monitoring infrastructures that capture performance relevant data during application execution assist in this task. The measured data forms the basis for identifying bottlenecks and optimizing the code.Monitoring infrastructures need mechanisms to record application activities in order to conduct measurements. Automatic instrumentation of the source code is the preferred method in most application scenarios. We introduce a plug-in for the LLVM infrastructure that enables automatic source code instrumentation at compile-time. In contrast to available instrumentation mechanisms in LLVM/Clang, our plug-in can selectively include/exclude individual application functions. This enables developers to fine-tune the measurement to the required level of detail while avoiding large runtime overheads due to excessive instrumentation.

Tschüter, Ronny↗