Search NASA⌕ Search

SEARCH · Search NASA

Results for “runtime prediction”

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 37 records · Page 2

Performance Model of MapReduce Iterative Applications for Hybrid Cloud Bursting

Hybrid cloud bursting (i.e., leasing temporary off-premise cloud resources to boost the overall capacity during peak utilization) can be a cost-effective way to deal with the increasing complexity of big data analytics, especially for iterative applications. However, the low throughput, high latency network link between the on-premise and off-premise resources (“weak link”) makes maintaining scalability difficult. While several data locality techniques have been designed for big data bursting on hybrid clouds, their effectiveness is difficult to estimate in advance. Yet such estimations are critical, because they help users decide whether the extra pay-as-you-go cost incurred by using the off-premise resources justifies the runtime speed-up. To this end, the current paper presents a performance model and methodology to estimate the runtime of iterative MapReduce applications in a hybrid cloud-bursting scenario. The paper focuses on the overhead incurred by the weak link at fine granularity, for both the map and the reduce phases. This approach enables high estimation accuracy, as demonstrated by extensive experiments at scale using a mix of real-world iterative MapReduce applications from standard big data benchmarking suites that cover a broad spectrum of data patterns. As a result, not only are the produced estimations accurate in absolute terms compared with experimental results, but they are also up to an order of magnitude more accurate than applying state-of-art estimation approaches originally designed for single-site MapReduce deployments.

97 MATHEMATICS AND COMPUTING↗

Many-Core Acceleration of a Discrete Ordinates Transport Mini-App at Extreme Scale, In: ISC High Performance 2016: High Performance Computing

Time-dependent deterministic discrete ordinates transport codes are an important class of application which provide significant challenges for large, many-core systems. One such challenge is the large memory capacity needed by the solve step, which requires us to have a scalable solution in order to have enough node-level memory to store all the data. In our previous work, we demonstrated the first implementation which showed a significant performance benefit for single node solves using GPUs. In this paper we extend our work to large problems and demonstrate the scalability of our solution on two Petascale GPU-based supercomputers: Titan at Oak Ridge and Piz Daint at CSCS. Our results show that our improved node-level parallelism scheme scales just as well across large systems as previous approaches when using the tried and tested KBA domain decomposition technique. We validate our results against an improved performance model which predicts the runtime of the main ‘sweep’ routine when running on different hardware, including CPUs or GPUs.

Computer Science↗

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↗

Task scheduling in dataflow computer architectures

Dataflow computers provide a platform for the solution of a large class of computational problems, which includes digital signal processing and image processing. Many typical applications are represented by a set of tasks which can be repetitively executed in parallel as specified by an associated dataflow graph. Research in this area aims to model these architectures, develop scheduling procedures, and predict the transient and steady state performance. Researchers at NASA have created a model and developed associated software tools which are capable of analyzing a dataflow graph and predicting its runtime performance under various resource and timing constraints. These models and tools were extended and used in this work. Experiments using these tools revealed certain properties of such graphs that require further study. Specifically, the transient behavior at the beginning of the execution of a graph can have a significant effect on the steady state performance. Transformation and retiming of the application algorithm and its initial conditions can produce a different transient behavior and consequently different steady state performance. The effect of such transformations on the resource requirements or under resource constraints requires extensive study. Task scheduling to obtain maximum performance (based on user-defined criteria), or to satisfy a set of resource constraints, can also be significantly affected by a transformation of the application algorithm. Since task scheduling is performed by heuristic algorithms, further research is needed to determine if new scheduling heuristics can be developed that can exploit such transformations. This work has provided the initial development for further long-term research efforts. A simulation tool was completed to provide insight into the transient and steady state execution of a dataflow graph. A set of scheduling algorithms was completed which can operate in conjunction with the modeling and performance tools previously developed. Initial studies on the performance of these algorithms were done to examine the effects of application algorithm transformations as measured by such quantities as number of processors, time between outputs, time between input and output, communication time, and memory size.

Katsinis, Constantine↗

Applying Machine Learning to Understand Write Performance of Large-scale Parallel Filesystems

In high-performance computing (HPC), I/O performance prediction offers the potential to improve the efficiency of scientific computing. In particular, accurate prediction can make runtime estimates more precise, guide users toward optimal checkpoint strategies, and better inform facility provisioning and scheduling policies. HPC I/O performance is notoriously difficult to predict and model, however, in large part because of inherent variability and a lack of transparency in the behaviors of constituent storage system components. In this work we seek to advance the state of the art in HPC I/O performance prediction by (1) modeling the mean performance to address high variability, (2) deriving model features from write patterns, system architecture and system configurations, and (3) employing Lasso regression model to improve model accuracy. We demonstrate the efficacy of our approach by applying it to a crucial subset of common HPC I/O motifs, namely, file-per-process checkpoint write workloads. We conduct experiments on two distinct production HPC platforms—Titan at the Oak Ridge Leadership Computing Facility and Cetus at the Argonne Leadership Computing Facility—to train and evaluate our models. We find that we can attain ≤ 30% relative error for 92.79% and 99.64% of the samples in our test set on these platforms, respectively.

Xie, Bing↗

MIC-SVM: Designing A Highly Efficient Support Vector Machine For Advanced Modern Multi-Core and Many-Core Architectures

Support Vector Machine (SVM) has been widely used in data-mining and Big Data applications as modern commercial databases start to attach an increasing importance to the analytic capabilities. In recent years, SVM was adapted to the field of High Performance Computing for power/performance prediction, auto-tuning, and runtime scheduling. However, even at the risk of losing prediction accuracy due to insufficient runtime information, researchers can only afford to apply offline model training to avoid significant runtime training overhead. To address the challenges above, we designed and implemented MICSVM, a highly efficient parallel SVM for x86 based multi-core and many core architectures, such as the Intel Ivy Bridge CPUs and Intel Xeon Phi coprocessor (MIC).

42 ENGINEERING↗

Aspen-based performance and energy modeling frameworks

Now with the anticipation of exascale architectures, energy consumption is becoming one of the critical design parameters, especially in light of the energy budget of 20-30 Megawatts set by the U.S. Department of Energy. Understanding an application’s execution pattern and its energy footprint is critical to improving application operation on a diverse heterogeneous architecture. Applying application-specific performance optimization can consequently improve energy consumption. However, this approach is only applicable to current systems. As we enter a new era of exascale architectures that is projected to contain more complex memory hierarchies, increased levels of parallelism, heterogeneity in hardware, and complex programming models and techniques, energy and performance management is getting more cumbersome. We therefore propose techniques that predict the energy consumption beforehand or at runtime to enable proactive tuning. Such energy prediction approaches must be generic and adapt themselves at runtime to changing application and hardware configurations. Most existing energy estimation and prediction approaches are empirical in nature and thus tied to current systems. To overcome this limitation, we propose two energy estimation techniques: ACEE (Algorithmic and Categorical Energy Estimation), which uses a combination of analytical and empirical modeling techniques; and AEEM (Aspen’s Embedded Energy Estimation), a system-level analytical energy estimation technique. Both of these models incorporate the Aspen domain specific language for performance modeling. We present the methodologies of these two models and test their accuracy using five proxy applications. We also describe three use cases.

79 ASTRONOMY AND ASTROPHYSICS↗

$\mathrm{PPT}$-Multicore: performance prediction of Open$\mathrm{MP}$ applications using reuse profiles and analytical modeling

In this report we present PPT-Multicore, an analytical model embedded in the Performance Prediction Toolkit (PPT) to predict parallel applications’ performance running on a multicore processor. PPT-Multicore builds upon our previous work towards a multicore cache model. We extract LLVM basic block labeled memory trace using an architecture-independent LLVM-based instrumentation tool only once in an application’s lifetime. The model uses the memory trace and other parameters from an instrumented sequentially executed binary. We use probabilistic and computationally efficient reuse profiles to predict the cache hit rates and runtimes of OpenMP programs’ parallel sections. We model Intel’s Broadwell, Haswell, and AMD’s Zen2 architectures and validate our framework using different applications from PolyBench and PARSEC benchmark suites. The results show that PPT-Multicore can predict cache hit rates with an overall average error rate of 1.23% while predicting the runtime with an error rate of 9.08%.

97 MATHEMATICS AND COMPUTING↗

Extending the Roofline Model for Asynchronous Many-Task Runtimes

New architectural trends and extreme scale parallelism challenge the efficient mapping of applications to large scale and emerging systems. Fine-grained Asynchronous Manytask Runtimes (AMR) offer a unique flexibility to hide the varying latencies of contentious operations, and facilitate the effective exploitation of abundant resources. By decomposing work into smaller asynchronous chunks, the impact of unpredictable operations can be mitigated through a tighter overlapping of tasks. Unfortunately, the cost of this decomposition comes at a potentially steep price in the form of runtime overhead which can at times rival the cost of the computation itself. This leads to the question, how large should a task be? A common practice for application developers is to experimentally determine the granularity of a task after a code has been parallelized. Instead, we propose a new methodology based on an extended Roofline model to provide practical upper bounds on the throughput performance of an application. First, we extend the Roofline model to support not only latency hiding analysis, but also a multidimensional amortized analysis. By combining this new methodology with a serial application and an AMR implementation, we can predict the worst case runtime overhead attribution of individual runtime features prior to the development of parallel code. Thus, this runtimecentric methodology can provide a vehicle for application/runtime codesign by providing a comprehensive bottleneck analysis based on existing runtime features.

Suetterlein, Joshua D.↗

Adaptive Impact-Driven Detection of Silent Data Corruption for HPC Applications

For exascale HPC applications, silent data corruption (SDC) is one of the most dangerous problems because there is no indication that there are errors during the execution. We propose an adaptive impact-driven method that can detect SDCs dynamically. The key contributions are threefold. (1) We carefully characterize 18 real-world HPC applications and discuss the runtime data features, as well as the impact of the SDCs on their execution results. (2) We propose an impact-driven detection model that does not blindly improve the prediction accuracy, but instead detects only influential SDCs to guarantee user-acceptable execution results. (3) Our solution can adapt to dynamic prediction errors based on local runtime data and can automatically tune detection ranges for guaranteeing low false alarms. Experiments show that our detector can detect 80-99.99% of SDCs with a false alarm rate less that 1% of iterations for most cases. The memory cost and detection overhead are reduced to 15% and 6.3%, respectively, for a large majority of applications.

Exascale HPC↗

Data-assisted combustion simulations with dynamic submodel assignment using random forests

This investigation outlines a data-assisted approach that employs random forest classifiers for local and dynamic submodel assignment in turbulent-combustion simulations. This method is demonstrated in simulations of a single-element GOX/GCH4 rocket combustor; a priori as well as a posteriori assessments are conducted to (i) evaluate the accuracy and adjustability of the classifier for targeting different quantities of interest (QoIs), and (ii) assess improvements, resulting from the data-assisted combustion model assignment, in predicting target QoIs during simulation runtime. Results from the a priori study show that random forests, trained with local flow properties as input variables and combustion model errors as training labels, assign three different combustion models – finite-rate chemistry (FRC), flamelet progress variable (FPV) model, and inert mixing (IM) – with reasonable classification performance even when targeting multiple QoIs. Applications in a posteriori studies demonstrate improved predictions from data-assisted simulations, in temperature and CO mass fraction, when compared with monolithic FPV calculations. An additional a posteriori data-assisted simulation of a modified configuration demonstrates that the present approach can be successfully applied to different configurations, as long as thermophysical behavior can be represented by the training data. Furthermore, these results demonstrate that this data-driven framework holds promise for dynamic combustion submodel assignments in reacting flow simulations.

42 ENGINEERING↗

An Efficient Silent Data Corruption Detection Method with Error-Feedback Control and Even Sampling for HPC Applications

The silent data corruption (SDC) problem is attracting more and more attentions because it is expected to have a great impact on exascale HPC applications. SDC faults are hazardous in that they pass unnoticed by hardware and can lead to wrong computation results. In this work, we formulate SDC detection as a runtime one-step-ahead prediction method, leveraging multiple linear prediction methods in order to improve the detection results. The contributions are twofold: (1) we propose an error feedback control model that can reduce the prediction errors for different linear prediction methods, and (2) we propose a spatial-data-based even-sampling method to minimize the detection overheads (including memory and computation cost). We implement our algorithms in the fault tolerance interface, a fault tolerance library with multiple checkpoint levels, such that users can conveniently protect their HPC applications against both SDC errors and fail-stop errors. We evaluate our approach by using large-scale traces from well-known, large-scale HPC applications, as well as by running those HPC applications on a real cluster environment. Experiments show that our error feedback control model can improve detection sensitivity by 34-189% for bit-flip memory errors injected with the bit positions in the range [20,30], without any degradation on detection accuracy. Furthermore, memory size can be reduced by 33% with our spatial-data even-sampling method, with only a slight and graceful degradation in the detection sensitivity.

Di, Sheng↗

Prometheus: Scalable and Accurate Emulation of Task-Based Applications on Many-Core Systems.

Modeling the performance of non-deterministic parallel applications on future many-core systems requires the development of novel simulation and emulation techniques and tools. We present “Prometheus”, a fast, accurate and modular emulation framework for task-based applications. By raising the level of abstraction and focusing on runtime synchronization, Prometheus can accurately predict applications’ performance on very large many-core systems. We validate our emulation framework against two real platforms (AMD Interlagos and Intel MIC) and report error rates generally below 4%. We, then, evaluate Prometheus’ performance and scalability: our results show that Prometheus can emulate a task-based application on a system with 512K cores in 11.5 hours. We present two test cases that show how Prometheus can be used to study the performance and behavior of systems that present some of the characteristics expected from exascale supercomputer nodes, such as active power management and processors with a high number of cores but reduced cache per core.

exascale systems↗

Extreme Heterogeneity 2018 - Productive Computational Science in the Era of Extreme Heterogeneity: Report for DOE ASCR Workshop on Extreme Heterogeneity

The 2018 Basic Research Needs Workshop on Extreme Heterogeneity identified five Priority Research Directions for realizing the capabilities needed to address the challenges posed in this era of rapid technological change. This report captures the outcomes of that workshop. In the context of extreme heterogeneity, it defines basic research needs and opportunities in computer science research to develop smart and trainable operating and runtime systems, programming environments, and predictive tools that will make future systems easier to adapt to scientists’ computing needs and easier for facilities to deploy securely.

97 MATHEMATICS AND COMPUTING↗

A Multi-Scale Inference, Estimation, and Prediction Engine for Earth System Modeling

We posit that AI methods can be leveraged to significantly enhance the predictive skill of forward Earth system modeling (ESM) activities. A hybrid framework incorporating traditional ESM modeling, inference methods, and AI techniques could make better use of both measured and computed information as well as computational resources by targeting inference tasks at program priorities, such as the predictability of precipitation extremes. This runtime pathway to closing the simulation/analysis-data/model improvement loop will streamline the traditional offline pathway to model improvement, which is based on domain science expertise, while suggesting guidance for further observations and measurements.

58 GEOSCIENCES↗

Mission Data System Java Edition Version 7

The Mission Data System framework defines closed-loop control system abstractions from State Analysis including interfaces for state variables, goals, estimators, and controllers that can be adapted to implement a goal-oriented control system. The framework further provides an execution environment that includes a goal scheduler, execution engine, and fault monitor that support the expression of goal network activity plans. Using these frameworks, adapters can build a goal-oriented control system where activity coordination is verified before execution begins (plan time), and continually during execution. Plan failures including violations of safety constraints expressed in the plan can be handled through automatic re-planning. This version optimizes a number of key interfaces and features to minimize dependencies, performance overhead, and improve reliability. Fault diagnosis and real-time projection capabilities are incorporated. This version enhances earlier versions primarily through optimizations and quality improvements that raise the technology readiness level. Goals explicitly constrain system states over explicit time intervals to eliminate ambiguity about intent, as compared to command-oriented control that only implies persistent intent until another command is sent. A goal network scheduling and verification process ensures that all goals in the plan are achievable before starting execution. Goal failures at runtime can be detected (including predicted failures) and handled by adapted response logic. Responses can include plan repairs (try an alternate tactic to achieve the same goal), goal shedding, ignoring the fault, cancelling the plan, or safing the system.

Reinholtz, William K.↗

Multiclass classification experiments

We seek a multiclass classifier that can satisfy the following requirements. 1. Predict among at least three classes. 2. Handle a moderately large number of correlated predictors. 3. Handle mixed categorical and continuous predictors. 4. Handle missing values (in some way). 5. Be trained with about n = 1,000 responses, and 6. Quantify prediction uncertainty. Based on 5. above, it is assumed that accuracy of prediction is paramount (relative to, say, runtime). All methods considered in this document take on the order of minutes (most take seconds) on a 2.8 GHz Quad-Core Intel Core i7 processor with 16GB RAM. Secondly, quantification of uncertainty is assumed to be a secondary task, since, with this small data size, resampling methods (i.e. bootstrapping) are possible. Further studies should quantify the feasibility and accuracy of resampling methods, and/or other methods, for quantifying uncertainty. R was used for all software.

97 MATHEMATICS AND COMPUTING↗

Improvements to Integrated Tradespace Analysis of Communications Architectures (ITACA) Network Loading Analysis Tool

NASA's SCENIC project aims to simplify and reduce the cost of space mission planning by replicating the analysis capabilities of commercially licensed software which are integrated with relevant analysis parameters specific to SCaN assets and SCaN supported user missions. SCENIC differs from current tools that perform similar analyses in that it 1) does not require any licensing fees, 2) will provide an all-in-one package for various analysis capabilities that normally requires add-ons or multiple tools to complete. As part of SCENIC's capabilities, the ITACA network loading analysis tool will be responsible for assessing the loading on a given network architecture and generating a network service schedule. ITACA will allow users to evaluate the quality of service of a given network architecture and determine whether or not the architecture will satisfy the mission's requirements. ITACA is currently under development, and the following improvements were made during the fall of 2017: optimization of runtime, augmentation of network asset pre-service configuration time, augmentation of Brent's method of root finding, augmentation of network asset FOV restrictions, augmentation of mission lifetimes, and the integration of a SCaN link budget calculation tool. The improvements resulted in (a) 25% reduction in runtime, (b) more accurate contact window predictions when compared to STK(Registered Trademark) contact window predictions, and (c) increased fidelity through the use of specific SCaN asset parameters.

analysis↗