Search NASA⌕ Search

SEARCH · Search NASA

Results for “HPC performance modeling”

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 721 records · Page 40

EReinit: Scalable and efficient fault-tolerance for bulk-synchronous MPI applications

Scientists from many different fields have been developing Bulk-Synchronous MPI applications to simulate and study a wide variety of scientific phenomena. Since failure rates are expected to increase in larger-scale future HPC systems, providing efficient fault-tolerance mechanisms for this class of applications is paramount. The global-restart model has been proposed to decrease the time of failure recovery in Bulk-Synchronous applications by allowing a fast reinitialization of MPI. However, the current implementations of this model have several drawbacks: they lack efficiency; their scalability have not been shown; and they require the use of the MPI profiling interface, which precludes the use of tools. Here, we present EReinit, an implementation of the global-restart model that addresses these problems. Our key idea and optimization is the co-design of basic fault-tolerance mechanisms such as failure detection, notification, and recovery between MPI and the resource manager in contrast to current approaches on which these mechanisms are implemented in MPI only. We demonstrate EReinit in three HPC programs and show that it is up to four times more efficient than existing solutions at 4,096 processes.

97 MATHEMATICS AND COMPUTING↗

Clacc: Translating OpenACC to OpenMP in Clang

OpenACC was launched in 2010 as a portable programming model for heterogeneous accelerators. Although various implementations already exist, no extensible, open-source, production-quality compiler support is available to the community. This deficiency poses a serious risk for HPC application developers targeting GPUs and other accelerators, and it limits experimentation and progress for the OpenACC specification. To address this deficiency, Clacc is a recent effort funded by the US Exascale Computing Project to develop production OpenACC compiler support for Clang and LLVM. A key feature of the Clacc design is to translate OpenACC to OpenMP to build on Clang's existing OpenMP compiler and runtime support. In this paper, we describe the Clacc goals and design. We also describe the challenges that we have encountered so far in our prototyping efforts, and we present some early performance results.

Denny, Joel↗

PAVE: An In Situ Framework for Scientific Visualization and Machine Learning Coupling

Machine learning (ML) has emerged as a tool for understanding data at scale. However, this new methodology comes at a cost because ML requires the use of even more HPC resources to generate ML algorithms. In addition to the compute resources required to develop ML algorithms, ML does not sidestep one of the biggest challenges on leading-edge HPC systems: the increasing gap between compute performance and I/O bandwidth. This has led to a strong push towards in situ, processing the data as it is generated, strategies to mitigate the I/O bottleneck. Unfortunately, there are no in situ frameworks dedicated to coupling scientific visualization and ML at scale to develop ML algorithms for scientific visualization. To address the ML and in situ visualization gap, we introduce PAVE. PAVE is an in situ framework which addresses the data management needs between visualisation and machine learning tasks. We demonstrate our framework with a case study that accelerates physically-based light rendering, path-tracing, through the use of a conditional Generative Adversarial neural Network (cGAN). PAVE couples the training over path-traced images resulting in a generative model able to produce scene renderings with accurate light transport and global illumination of a quality comparable to offline approaches in a more efficient manner.

Kim, Mark B.↗

Dynamic Undervolting to Improve Energy Efficiency on Multicore X86 CPUs

Chip manufacturers introduce redundancy at various levels of CPU design to guarantee correct operation, even for worst-case combinations of non-idealities in process variation and system operating conditions. This redundancy is implemented partly in the form of voltage margins. However, for a wide range of real-world execution scenarios these margins are excessive and merely translate to increased power consumption, hindering the effort towards higher-energy efficiency in both HPC and general purpose computing. Our study on the x86-64 Haswell and Skylake multicore microarchitectures reveals-wide voltage margins, which vary across different microarchitectures, different chip parts of the same microarchitecture, and across different workloads. We find that it is necessary to quantify-voltage margins using multi-threaded and multi-instance workloads, as characterization with single-threaded and single-instance workloads that do not stress the CPU to its full capacity typically identifies overly optimistic margins that lead to errors when applied in realistic program execution scenarios. In addition, we introduce, deploy and evaluate a run-time governor that dynamically reduces the supply voltage of modern multicore x86-64 CPUs. Our governor employs a model that takes as input a set of performance metrics which are directly measurable via performance monitoring counters and have high predictive value for the minimum tolerable supply voltage (V min ), to predict and apply the appropriate reduction for the workload at hand. Compared with the conventional DVFS governor, our approach in this study achieves up to 42 percent energy savings for the Skylake family and 34 percent for the Haswell family for complex, real-world applications.

97 MATHEMATICS AND COMPUTING↗

The Livermore Brain: Massive Deep Learning Networks Enabled by High Performance Computing

The proliferation of inexpensive sensor technologies like the ubiquitous digital image sensors has resulted in the collection and sharing of vast amounts of unsorted and unexploited raw data. Companies and governments who are able to collect and make sense of large datasets to help them make better decisions more rapidly will have a competitive advantage in the information era. Machine Learning technologies play a critical role for automating the data understanding process; however, to be maximally effective, useful intermediate representations of the data are required. These representations or “features” are transformations of the raw data into a form where patterns are more easily recognized. Recent breakthroughs in Deep Learning have made it possible to learn these features from large amounts of labeled data. The focus of this project is to develop and extend Deep Learning algorithms for learning features from vast amounts of unlabeled data and to develop the HPC neural network training platform to support the training of massive network models. This LDRD project succeeded in developing new unsupervised feature learning algorithms for images and video and created a scalable neural network training toolkit for HPC. Additionally, this LDRD helped create the world’s largest freely-available image and video dataset supporting open multimedia research and used this dataset for training our deep neural networks. This research helped LLNL capture several work-for-others (WFO) projects, attract new talent, and establish collaborations with leading academic and commercial partners. Finally, this project demonstrated the successful training of the largest unsupervised image neural network using HPC resources and helped establish LLNL leadership at the intersection of Machine Learning and HPC research.

97 MATHEMATICS AND COMPUTING↗

Scalable Energy Efficiency with Resilience for High Performance Computing Systems: A Quantitative Methodology

Energy efficiency and resilience are two crucial challenges for HPC systems to reach exascale. While energy efficiency and resilience issues have been extensively studied individually, little has been done to understand the interplay between energy efficiency and resilience for HPC systems. Decreasing the supply voltage associated with a given operating frequency for processors and other CMOS-based components can significantly reduce power consumption. However, this often raises system failure rates and consequently increases application execution time. In this work, we present an energy saving undervolting approach that leverages the mainstream resilience techniques to tolerate the increased failures caused by undervolting.

energy, resilience↗

MARS: Malleable Actor-Critic Reinforcement Learning Scheduler

In this paper, we introduce MARS, a new scheduling system for HPC-cloud infrastructures based on a cost-aware, flexible reinforcement learning approach, which serves as an intermediate layer for next generation HPC-cloud resource manager. MARS ensembles the pre-trained models from heuristic workloads and decides on the most cost-effective strategy for optimization. A whole workflow application would be split into several optimizable dependent sub-tasks, then based on the pre- defined resource management plan, a reward will be generated after executing a scheduled task. Lastly, MARS updates the Deep Neural Network (DNN) model based on the reward. MARS is designed to optimize the existing models through reinforcement mechanisms. MARS adapts to the dynamics of workflow applications, selects the most cost-effective scheduling solution among pre-built scheduling strategies (backfilling, SJF, etc.) and self- learning deep neural network model at run-time. We evaluate MARS with different real-world workflow traces. MARS can achieve 5%-60% increased performance compare to state-of-the- art approaches.

Baheri, Betis↗

Thermal Management for FPGA Nodes in HPC Systems

The integration of FPGAs into large-scale computing systems is gaining attention. In these systems, real-time data handling for networking, tasks for scientific computing, and machine learning can be executed with customized datapaths on reconfigurable fabric within heterogeneous compute nodes. At the same time, thermal management, particularly battling the cooling cost and guaranteeing the reliability, is a continuing concern. The introduction of new heterogeneous components into HPC nodes only adds further complexities to thermal modeling and management. The thermal behavior of multi-FPGA systems deployed within large compute clusters is less explored. Here, we first show that the thermal behaviors of different FPGAs of the same generation can vary due to their physical locations in a rack and process variation, even though they are running the same tasks. We present a machine learning–based model to capture the thermal behavior of each individual FPGA in the cluster. We then propose two thermal management strategies guided by our thermal model. First, we mitigate thermal variation and hotspots across the cluster by proactive thermal-aware task placement. Under the tested system and benchmarks, we achieve up to 26.4° C and on average 13.3° C system temperature reduction with no performance penalty. Second, we utilize this thermal model to guide HLS parameter tuning at the task design stage to achieve improved thermal response after deployment.

97 MATHEMATICS AND COMPUTING↗

The Influence of Computer Architecture on Performance and Scaling for Hypersonic Flow Simulations

It is critical to understand how hypersonic simulation tools perform on a range of computational platforms. This information will aid in the acquisition of appropriate hardware and the potential refactoring of hypersonic codes to run on different systems. In this paper, we consider two representative high-speed reacting flow cases: a model Mach 8 hypersonic waverider glide vehicle and a model hydrocarbon-fueled hypersonic ramjet propulsion system. In both scenarios, the flow fields are in chemical non-equilibrium and are modeled by the multi-species reacting Navier-Stokes equations. For these simulations we use several hypersonic simulation tools, including US3D, Kestrel, FUN3D, and the JENRER© flow solver. We explore several high performance computing systems containing IntelR© XeonR© Platinum processors, AMD EPYCTM7702 processors, and NVIDIAR© Tesla V100 devices. We compare performance and strong scaling between the different systems.

CPU↗

SIRIUS: Enabling Progressive Data Exploration for Extreme-Scale Scientific Data

Scientific simulations on high performance computing (HPC) platforms generate large quantities of data. To bridge the widening gap between compute and I/O, and enable data to be more efficiently stored and analyzed, simulation outputs need to be refactored, reduced, and appropriately mapped to storage tiers. However, a systematic solution to support these steps has been lacking in the current HPC software ecosystem. To that end, this paper develops SIRIUS, a progressive JPEG-like data management scheme for storing and analyzing big scientific data. It co-designs data decimation, compression, and data storage, taking the hardware characteristics of each storage tier into considerations. With reasonably low overhead, our approach refactors simulation data, using either topological or uniform decimation, into a much smaller, reduced-accuracy base dataset, and a series of deltas that is used to augment the accuracy if needed. The base dataset and deltas are compressed and written to multiple storage tiers. Data saved on different tiers can then be selectively retrieved to restore the level of accuracy that satisfies data analytics. Thus, SIRIUS provides a paradigm shift towards elastic data analytics and enables end users to make trade-offs between analysis speed and accuracy on-the-fly. This paper further develops algorithms to preserve statistics for data decimation, a common requirement for reducing data. Here, we assess the impact of SIRIUS on unstructured triangular meshes, a pervasive data model used in scientific simulations. In particular, we evaluate two realistic use cases: the blob detection in fusion and high-pressure area extraction in computational fluid dynamics.

97 MATHEMATICS AND COMPUTING↗

HPC Interconnection Networks: The Key to Exascale Computing

Scientists from many domains desire to address problems within the next decade that, by all estimates, require computer systems that can achieve sustained exaflop computing rates (i.e., 1 x 10{sup 18} floating point operations per second) with real-world applications. Simply scaling existing designs is insufficient: analysis of current technological trends suggests that only a few architectural components are on track to reach the performance levels needed for exascale computing. The network connecting computer system nodes presents a particularly difficult challenge because of the prevalence of a wide variety of communication patterns and collective communication operations in algorithms used in scientific applications and their tendency to be the most significant limit to application scalability. Researchers at Oak Ridge National Laboratory and elsewhere are actively working to overcome these network-related scalability barriers using advanced hardware and software design, alternative network topologies, and performance prediction using modeling and simulation.

97 MATHEMATICS AND COMPUTING↗

ASCR Cybersecurity for Scientific Computing Integrity - Research Pathways and Ideas Workshop

At the request of the U.S. Department of Energy's (DOE) Office of Science (SC) Advanced Scientific Computing Research (ASCR) program office, a workshop was held June 2-3, 2015, in Gaithersburg, MD, to identify potential long term (10 to +20 year) cybersecurity fundamental basic research and development challenges, strategies and roadmap facing future high performance computing (HPC), networks, data centers, and extreme-scale scientific user facilities. This workshop was a follow-on to the workshop held January 7-9, 2015, in Rockville, MD, that examined higher level ideas about scientific computing integrity specific to the mission of the DOE Office of Science. Issues included research computation and simulation that takes place on ASCR computing facilities and networks, as well as network-connected scientific instruments, such as those run by various DOE Office of Science programs. Workshop participants included researchers and operational staff from DOE national laboratories, as well as academic researchers and industry experts. Participants were selected based on the submission of abstracts relating to the topics discussed in the previous workshop report [1] and also from other ASCR reports, including "Abstract Machine Models and Proxy Architectures for Exascale Computing" [27], the DOE "Preliminary Conceptual Design for an Exascale Computing Initiative" [28], and the January 2015 machine learning workshop [29]. The workshop was also attended by several observers from DOE and other government agencies. The workshop was divided into three topic areas: (1) Trustworthy Supercomputing, (2) Extreme-Scale Data, Knowledge, and Analytics for Understanding and Improving Cybersecurity, and (3) Trust within High-end Networking and Data Centers. Participants were divided into three corresponding teams based on the category of their abstracts. The workshop began with a series of talks from the program manager and workshop chair, followed by the leaders for each of the three topics and a representative of each of the four major DOE Office of Science Advanced Scientific Computing Research Facilities: the Argonne Leadership Computing Facility (ALCF), the Energy Sciences Network (ESnet), the National Energy Research Scientific Computing Center (NERSC), and the Oak Ridge Leadership Computing Facility (OLCF). The rest of the workshop consisted of topical breakout discussions and focused writing periods that produced much of this report.

97 MATHEMATICS AND COMPUTING↗

A Knowledge Graph Approach for the Secondary Use of Cancer Registry Data

Population-based central cancer registries collect valuable structured and unstructured data primarily for cancer surveillance and research, enhancing insights into clinical features associated with cancer occurrence, cancer treatment, and cancer outcomes to guide interventions which reduce the cancer burden. Cancer registries primarily collect data on (1) cancer type (case or tumor); (2) patient demographics such as age, gender, and residential address at time of diagnosis; (3) planned first course of treatment; and (4) date of last contact, vital status, and cause of death. Cancer registry data is dynamic, structured data, which is extracted from many unstructured sources such as electronic healthcare records, and consolidated for reporting and other purposes. While available advanced analytic tools such as SEER*Stat have the ability to build SAS queries, we, however, explore an innovative knowledge graph approach to organizing cancer registry data for advanced analytics and visualization, which has unique advantages over approaches of existing tools. This innovative knowledge graph approach semantically enriches the data and easily enables linkage with third-party data, which can better explain variation in outcomes. We have developed a prototype knowledge graph based on data from the Louisiana Tumor Registry and other publicly available datasets including Behavioral Risk Factor Surveillance System, Clinical Trials, DBpedia, GeoNames, Rural-Urban Continuum Codes, and Semantic MEDLINE. The resource description framework (RDF) data model was selected to represent our knowledge graph, which contains more than 25 billion triples and is ~4TB in storage size. To exhibit the benefits of the knowledge graph approach, we used scenario specific queries, which find the relationships between cancer treatment sequences and outcomes. To illustrate its ease of use in iterative analysis, the knowledge graph was linked to external datasets for performing complex queries across multiple datasets. In addition, we used knowledge graphs to identify data discrepancies and to handle schema changes. Finally, we visualized the knowledge graph to discover data patterns. Our results demonstrate this graph-based solution enables cancer researchers to execute complex queries and more easily perform iterative analyses to improve understanding of cancer registry data. In the future, we would like to use high-performance computing (HPC) resources for faster-generating hypotheses with clinical potential from our knowledge graph.

Hasan, S M Shamimul↗

A compute-bound formulation of Galerkin model reduction for linear time-invariant dynamical systems

This work aims to advance computational methods for projection-based reduced-order models (ROMs) of linear time-invariant (LTI) dynamical systems. For such systems, current practice relies on ROM formulations expressing the state as a rank-1 tensor (i.e., a vector), leading to computational kernels that are memory bandwidth bound and, therefore, ill-suited for scalable performance on modern architectures. This weakness can be particularly limiting when tackling many-query studies, where one needs to run a large number of simulations. This work introduces a reformulation, called rank-2 Galerkin, of the Galerkin ROM for LTI dynamical systems which converts the nature of the ROM problem from memory bandwidth to compute bound. We present the details of the formulation and its implementation, and demonstrate its utility through numerical experiments using, as a test case, the simulation of elastic seismic shear waves in an axisymmetric domain. We quantify and analyze performance and scaling results for varying numbers of threads and problem sizes. In conclusion, we present an end-to-end demonstration of using the rank-2 Galerkin ROM for a Monte Carlo sampling study. We show that the rank-2 Galerkin ROM is one order of magnitude more efficient than the rank-1 Galerkin ROM (the current practice) and about 970 times more efficient than the full-order model, while maintaining accuracy in both the mean and statistics of the field.

97 MATHEMATICS AND COMPUTING↗

Lamellar: A Rust-based Asynchronous Tasking and PGAS Runtime for High Performance Computing

Cybersecurity is one of the largest concerns in modern computing, impacting and dictating how governments, private corporations, and individuals interact with and live in an increasingly digital world. The NSA has recently released a memo [ 1] on “Software Memory Safety” where they highlight that both Microsoft and Google have stated around 70% of software vulnerabilities were due to memory safety issues. Although languages such as C and C++ provide freedom and flexibility with memory management, guaran- teeing safety falls mostly on the developer. The NSA recommends using “memory safe” languages whenever possible. In this paper we introduce Lamellar, an asynchronous tasking and PGAS HPC runtime written in Rust, one such "memory safe" language. We describe the entire Lamellar stack, from network interfaces to high- level abstractions such as distributed LamellarArrays and Active Messages. We conclude by showing comparable performance to legacy PGAS runtimes (e.g. OpenSHMEM) on a subset of the BALE kernel suite while maintaining strong memory safety principles.

HPC Software Systems, Rust Programming Language, P↗

Large-Scale Materials Modeling at Quantum Accuracy: Ab Initio Simulations of Quasicrystals and Interacting Extended Defects in Metallic Alloys

Ab initio electronic-structure has remained dichotomous between achievable accuracy and length-scale. Quantum many-body (QMB) methods realize quantum accuracy but fail to scale. Density functional theory (DFT) scales favorably but remains far from quantum accuracy. We present a framework that breaks this dichotomy by use of three interconnected modules: (i) invDFT: a methodological advance in inverse DFT linking QMB methods to DFT; (ii) MLXC: a machine-learned density functional trained with invDFT data, commensurate with quantum accuracy; (iii) DFT-FE-MLXC: an adaptive higher-order spectral finite-element (FE) based DFT implementation that integrates MLXC with efficient solver strategies and HPC innovations in FE-specific dense linear algebra, mixed-precision algorithms, and asynchronous compute-communication. Furthermore, we demonstrate a paradigm shift in DFT that not only provides an accuracy commensurate with QMB methods in ground-state energies, but also attains an unprecedented performance of 659.7 PFLOPS (43.1% peak FP64 performance) on 619,124 electrons using 8,000 GPU nodes of Frontier supercomputer.

density functional theory↗

Failure prediction using machine learning in a virtualised HPC system and application

Failure is an increasingly important issue in high performance computing and cloud systems. As large-scale systems continue to grow in scale and complexity, mitigating the impact of failure and providing accurate predictions with sufficient lead time remains a challenging research problem. Traditional existing fault-tolerance strategies such as regular check-pointing and replication are not adequate because of the emerging complexities of high performance computing systems. This necessitates the importance of having an effective as well as proactive failure management approach in place aimed at minimizing the effect of failure within the system. With the advent of machine learning techniques, the ability to learn from past information to predict future pattern of behaviours makes it possible to predict potential system failure more accurately. Thus, in this paper, we explore the predictive abilities of machine learning by applying a number of algorithms to improve the accuracy of failure prediction. We have developed a failure prediction model using time series and machine learning, and performed comparison based tests on the prediction accuracy. The primary algorithms we considered are the support vector machine (SVM), random forest (RF), k-nearest neighbors (KNN), classification and regression trees (CART) and linear discriminant analysis (LDA). Experimental results indicates that the average prediction accuracy of our model using SVM when predicting failure is 90% accurate and effective compared to other algorithms. This finding implies that our method can effectively predict all possible future system and application failures within the system.

Mohammed, Bashir↗

W17_geowave “3D full waveform geophysical models”

Performance of the MCMC inversion according to the number of cores for the computation. A) 64 cores. B) 480 cores. C) 816 cores. The true model is represented by the black line. Vsv is the wave speed of S waves polarized in the vertical plane, ξ is an anisotropy parameter. The Earth is highly anisotropics; the wavespeed of seismic waves depends on the polarization of the wave. Seismic inversion of the elastic structure is usually limited to isotropic information such as Vsv. Our research looked at the inversion of Earth anisotropy.

58 GEOSCIENCES↗