Summary With the rise of exascale systems and large, data‐centric workflows, the need to observe and analyze high performance computing (HPC) applications during their execution is becoming increasingly important. HPC applications are typically not designed with online monitoring in mind, therefore, the observability challenge lies in being able to access and analyze interesting events with low overhead while seamlessly integrating such capabilities into existing and new applications. We explore how our service‐based observation, monitoring, and analytics (SOMA) approach to collecting and aggregating both application‐specific diagnostic data and performance data addresses these needs. We present our SOMA framework and demonstrate its viability with LULESH, a hydrodynamics proxy application. Then we focus on Astaroth, a multi‐GPU library for stencil computations, highlighting the integration of the TAU and APEX performance tools and SOMA for application and performance data monitoring.

Understanding the behavior of software in execution is a key step in identifying and fixing performance issues. This is especially important in high performance computing contexts where even minor performance tweaks can translate into large savings in terms of computational resource use. To aid performance analysis, developers may collect an execution trace-a chronological log of program activity during execution. As traces represent the full history, developers can discover a wide array of possibly previously unknown performance issues, making them an important artifact for exploratory performance analysis. However, interactive trace visualization is difficult due to issues of data size and complexity of meaning. Traces represent nanosecond-level events across many parallel processes, meaning the collected data is often large and difficult to explore. The rise of asynchronous task parallel programming paradigms complicates the relation between events and their probable cause. To address these challenges, we conduct a continuing design study in collaboration with high performance computing researchers. We develop diverse and hierarchical ways to navigate and represent execution trace data in support of their trace analysis tasks. Through an iterative design process, we developed Traveler, an integrated visualization platform for task parallel traces. Traveler provides multiple linked interfaces to help navigate trace data from multiple contexts. We evaluate the utility of Traveler through feedback from users and a case study, finding that integrating multiple modes of navigation in our design supported performance analysis tasks and led to the discovery of previously unknown behavior in a distributed array library.

The current practice for fusion scientists running first principle simulations on high performance computing plat-forms is to either run their simulations and output their data for post-hoc analysis, or to place in situ analytics into their code. In this paper we examine a complex workflow using XGC fusions simulation run on the Oak Ridge Leadership Computing Facility's supercomputer Summit, which also involve three anal-yses as part of the results necessary for scientific discovery. We discuss the challenges faced when implementing these algorithms and present an original hybrid staging technique to help enable the physicists to make discoveries during the execution of the simulation. By creating this infrastructure, we can examine complicated physics results, which may not have been possible without the infrastructure. For example, our work enables the online visualization of turbulent homoclinic tangle around the magnetic X-point, breaking the last confinement surface. This visualization could help fusion scientists to better understand and improve the turbulence spread of plasma exhaust heat, which is crucial toward realizing plasmas beyond the currently accessible physics regimes of present-day tokamak reactors. The physics of turbulent homoclinic tangle will be reported in a future physics publication, by utilizing the original online analysis/visualization framework presented in this paper.

Abstract Science applications preparing for the exascale era are increasingly exploring in situ computations comprising of simulation‐analysis‐reduction pipelines coupled in‐memory. Efficient composition and execution of such complex pipelines for a target platform is a codesign process that evaluates the impact and tradeoffs of various application‐ and system‐specific parameters. In this article, we describe a toolset for automating performance studies of composed HPC applications that perform online data reduction and analysis. We describe Cheetah, a new framework for composing parametric studies on coupled applications, and Savanna, a runtime engine for orchestrating and executing campaigns of codesign experiments. This toolset facilitates understanding the impact of various factors such as process placement, synchronicity of algorithms, and storage versus compute requirements for online analysis of large data. Ultimately, we aim to create a catalog of performance results that can help scientists understand tradeoffs when designing next‐generation simulations that make use of online processing techniques. We illustrate the design of Cheetah and Savanna, and present application examples that use this framework to conduct codesign studies on small clusters as well as leadership class supercomputers.

Octo-Tiger is a code for modeling three-dimensional self-gravitating astrophysical fluids. It was particularly designed for the study of dynamical mass transfer between interacting binary stars. Octo-Tiger is parallelized for distributed systems using the asynchronous many-task runtime system, the C++ standard library for parallelism and concurrency (HPX) and utilizes CUDA for its gravity solver. Recently, we have remodeled Octo-Tiger’s hydro solver to use a three-dimensional reconstruction scheme. In addition, we have ported the hydro solver to GPU using CUDA kernels. We present scaling results for the new hydro kernels on ORNL’s Summit machine using a Sedov-Taylor blast wave problem. We also compare Octo-Tiger’s new hydro scheme with its old hydro scheme, using a rotating star as a test problem.

High-performance computing (HPC) software is evolving to support an increasingly diverse set of applications and heterogeneous hardware architectures. As part of this evolution, the construction of scientific software has shifted from a traditional monolithic message passing interface executable model to a coupled, services-style model in which simulations run alongside a host of distributed HPC data services within the same batch job allocation. Microservices have emerged as a powerful new way to build these distributed data services through a composition model. However, performance analysis of composed microservices is a daunting challenge. It requires collecting, monitoring, aggre-gating, and exporting performance data from multiple sources. To be effective, the design of such a monitoring solution must allow for seamless integration into HPC applications and distributed services alike, be scalable, operate with a low overhead, and take advantage of the HPC platform. We propose SYMBIOMON, a monitoring service that is built by composing high-performance microservices. We describe its design and implementation within the context of the Mochi framework. SYMBIOMON combines a time-series data model with existing Mochi data services to collect, aggregate, and export performance metrics in a distributed manner. SYMBIOMON enables seamless, low-overhead monitoring and analysis of data services and HPC applications alike. Using HEPnOS, a production-quality Mochi data service, we demonstrate the use of SYMBIOMON to identify better service configurations.

Part of the promise of exascale computing and the next generation of scientific simulation codes is the ability to bring together time and spatial scales that have traditionally been treated separately. This enables creating complex coupled simulations and in situ analysis pipelines, encompassing such things as "whole device" fusion models or the simulation of cities from sewers to rooftops. Unfortunately, the HPC analysis tools that have been built up over the preceding decades are ill suited to the debugging and performance analysis of such computational ensembles. In this paper, we present a new vision for performance measurement and understanding of HPC codes, MonitoringAnalytics (MONA). MONA is designed to be a flexible, high performance monitoring infrastructure that can perform monitoring analysis in place or in transit by embedding analytics and characterization directly into the data stream, without relying upon delivering all monitoring information to a central database for post-processing. It addresses the trade-offs between the prohibitively expensive capture of all performance characteristics and not capturing enough to detect the features of interest. We demonstrate several uses of MONA; capturing and indexing multi-executable performance profiles to enable later processing, extraction of performance primitives to enable the generation of customizable benchmarks and performance skeletons, and extracting communication and application behaviors to enable better control and placement for the current and future runs of the science ensemble. Relevant performance information based on a system for MONA built from ADIOS and SOSflow technologies is provided for DOE science applications and leadership machines.

Several robust performance systems have been created for parallel machines with the ability to observe diverse aspects of application execution on different hardware platforms. All of these are designed with the objective to support measurement methods that are efficient, portable, and scalable. For these reasons, the performance measurement infrastructure is tightly embedded with the application code and runtime execution environment. As parallel software and systems evolve, especially towards more heterogeneous, asynchronous, and dynamic operation, it is expected that the requirements for performance observation and awareness will change. For instance, heterogeneous machines introduce new types of performance data to capture and performance behaviors to characterize. Furthermore, there is a growing interest in interacting with the performance infrastructure for in situ analytics and policy-based control. The problem is that an existing performance system architecture could be constrained in its ability to evolve to meet these new requirements. The paper reports our research efforts to address this concern in the context of the TAU Performance System. In particular, we consider the use of a powerful plugin model to both capture existing capabilities in TAU and to extend its functionality in ways it was not necessarily conceived originally. The TAU plugin architecture supports three types of plugin paradigms: EVENT, TRIGGER, and AGENT. We demonstrate how each operates under several different scenarios. Results from larger-scale experiments are shown to highlight the fact that efficiency and robustness can be maintained, while new flexibility and programmability can be offered that leverages the power of the core TAU system while allowing significant and compelling extensions to be realized.

As the era of high frequency, single core processors have come to a close, the new paradigm of many core processors has come to dominate. In response to these systems, asynchronous multitasking runtime systems have been developed as a promising solution to efficiently utilize these newly available hardware. Asynchronous multitasking runtime systems work by dividing a problem into a large number of fine grained tasks. However, as the number of tasks created increase, the overheads associated with task creation and management cannot be ignored. Task inlining, a method where the parent thread consumes a child thread, enables the runtime system to achieve the balance between parallelism and its overhead. As largely impacted by different processor architectures, the decision of task inlining is dynamic in nature. In this research, we present adaptive techniques for deciding, at runtime, whether a particular task should be inlined or not. We present two policies, a baseline policy that makes inlining decision based on a fixed threshold and an adaptive policy which decides the threshold dynamically at runtime. We also evaluate and justify the performance of these policies on different processor architectures. To the best of our knowledge, this is the first study of the impacts of adaptive policy at runtime for task inlining in an asynchronous multitasking runtime system on different processor architectures. From experimentation, we find that the baseline policy improves the execution time from 7.61% to 54.09%. Furthermore, the adaptive policy improves over the baseline policy by up to 74%.

We study the simulation of stellar mergers, which requires complex simulations with high computational demands. We have developed Octo-Tiger, a finite volume grid-based hydrodynamics simulation code with Adaptive Mesh Refinement which is unique in conserving both linear and angular momentum to machine precision. To face the challenge of increasingly complex, diverse, and heterogeneous HPC systems, Octo-Tiger relies on high-level programming abstractions.

With the growing computational complexity of science and the complexity of new and emerging hardware, it is time to re-evaluate the traditional monolithic design of computational codes. One new paradigm is constructing larger scientific computational experiments from the coupling of multiple individual scientific applications, each targeting their own physics, characteristic lengths, and/or scales. We present a framework constructed by leveraging capabilities such as in-memory communications, workflow scheduling on HPC resources, and continuous performance monitoring. This code coupling capability is demonstrated by a fusion science scenario, where differences between the plasma at the edges and at the core of a device have different physical descriptions. This infrastructure not only enables the coupling of the physics components, but it also connects in situ or online analysis, compression, and visualization that accelerate the time between a run and the analysis of the science content. Results from runs on Titan and Cori are presented as a demonstration.

The Stingray raytracer was developed for marine seismology to compute minimum travel time from all sources in an earth model to determine the 3D geophysical structure below the ocean floor. The original sequential implementation of Stingray used Dijkstra's single-source, shortest-path (SSSP) algorithm. A data parallel version of Stingray was developed based on the Bellman-Ford-Moore iterative SSSP algorithm. Single node experiments demonstrated performance improvements from parallelization with multicore (using OpenMP) and manycore processors (using CUDA). Calculating seismic ray paths for larger earth models requires distributed, multi-node algorithms utilizing domain decomposition methods. Preliminary 2D decomposition strategies show promising scaling results. However, a general 3D decomposition methodology is needed to handle any seismic raytracing problem on any HPC computing platform. In this paper, we present Stingray-HPC, a framework for scalable seismic raytracing which can automatically decompose a 3D earth model across nodes in a distributed environment, allocate ghost cell regions for iterative updates, coordinate ghost cell communications, and test for global convergence. Stingray-HPC is implemented with MPI and either OpenMP or CUDA for node- level calculations. Our results validate Stingray-HPC's ability to handle large models (over a billion points) and to solve these models efficiently at scale up to 512 GPU nodes.

As the memory and storage hierarchy get deeper and more complex, it is important to have new benchmarks and evaluation tools that allow us to explore the emerging middleware solutions to use this hierarchy. Skel is a tool aimed at automating and refining this process of studying HPC I/O performance. It works by generating application I/O kernel/benchmarks as determined by a domain-specific model. This paper provides some techniques for extending Skel to address new situations and to answer new research questions. For example, we document use cases as diverse as using Skel to troubleshoot I/O performance issues for remote users, refining an I/O system model, and facilitating the development and testing of a mechanism for runtime monitoring and performance analytics. We also discuss data oriented extensions to Skel to support the study of compression techniques for Exascale scientific data management.

Power is the most critical resource for the exascale high performance computing. In the future, system administrators might have to pay attention to the power consumption of the machine under different work loads. Hence, each application may have to run with an allocated power budget. Thus, achieving the best performance on future machines requires optimal performance subject to a power constraint. This additional performance requirement should not be the responsibility of HPC~(High Performance Computing) application developers. Optimizing the performance for a given power budget should be the responsibility of high-performance system software stack. Modern machines allow power capping of CPU and memory to implement power budgeting strategy. Finding the best runtime environment for a node at a given power level is important to get the best performance. This paper presents ARCS (Adaptive Runtime Configuration Selection) frameworkthat automatically selects the best runtime configuration for each OpenMPparallel region at a given power level. The framework uses OMPT (OpenMP Tools) API, APEX(Autonomic Performance Environment for eXascale), and Active Harmony frameworksto explore configuration search space and selects the best number of threads, scheduling policy, and chunk size for a given power level at run-time. We test ARCS using the NAS Parallel Benchmark, and proxy application LULESH with Intel Sandybridge, and IBM Power multi-core architectures. We show that for a given power level, efficient OpenMP runtime parameter selection can improve the execution time and energy consumption of an application up to 40% and 42% respectively.

A hybrid parallel measurement system offers the potential to fuse the principal advantages of probe-based tools, with their exact measures of performance and ability to capture event semantics, and sampling-based tools, with their ability to observe performance detail with less overhead. Creating a hybrid profiling solution is challenging because it requires new mechanisms for integrating probe and sample measurements and calculating profile statistics during execution. In this paper, we describe a general hybrid parallel profiling tool that has been implemented in the TAU Performance System. Its generality comes from the fact that all of the features of the individual methods are retained and can be flexibly controlled when combined to address the measurement requirements for a particular parallel application. The design of the hybrid profiling approach is described and the implementation of the prototype in TAU presented. We demonstrate hybrid profiling functionality first on a simple sequential program and then show its use for several OpenMP parallel codes from the NAS Parallel Benchmark. These experiments also highlight the improvements in overhead efficiency made possible by hybrid profiling. A large-scale ocean modeling code based on OpenMP and MPI, MPAS-Ocean, is used to show how the TAU hybrid profiling tool can be effective at exposing performance-limiting behavior that would be difficult to identify otherwise.

The emergence of Petascale systems has raised new challenges to performance analysis tools. Understanding every single detail of an execution is important to bridge the gap between the theoretical peak and the actual performance achieved. Tracing tools are the best option when it comes to providing detailed information about the application behavior, but not without liabilities. The amount of information that a single execution can generate grows so fast that it easily becomes unmanageable. An effective analysis in such scenarios necessitates the intelligent selection of information. In this paper we present an on-line performance tool based on spectral analysis of signals that automatically identifies the different computing phases of the application as it runs, selects a few representative periods and decides the granularity of the information gathered for these regions. As a result, the execution is completely characterized at different levels of detail, reducing the amount of data collected while maximizing the amount of useful information presented for the analysis.

As access to supercomputing resources is becoming more and more commonplace, performance analysis tools are gaining importance in order to decrease the gap between the application performance and the supercomputers' peak performance. Performance analysis tools allow the analyst to understand the idiosyncrasies of an application in order to improve it. However, these tools require monitoring regions of the application to provide information to the analysts, leaving non-monitored regions of code unknown, which may result in lack of understanding of important regions of the application. In this paper we describe an automated methodology that reports very detailed application insights and improves the analysis experience of performance tools based on traces. We apply this methodology to three production applications and provide suggestions on how to improve their performance. Our methodology uses computation burst clustering and a mechanism called folding. While clustering automatically detects application structure, folding combines instrumentation and sampling to augment the performance analysis details. Folding provides fine grain performance information from coarse grain sampling on iterative applications. Folding results closely resemble the performance data gathered from fine grain sampling with an absolute mean difference less than 5% without overhead of fine grain.

Balancing the workload in parallel applications is a difficult task, even in conventional cases. Many computing cycles are wasted when the load is not evenly balanced across processing nodes. Global load balance analysis may determine that an application is well balanced, when in fact the application has hidden inefficiencies. In this paper, we consider the load balance of parallel applications which present unique challenges in the analysis process. We have performed trace analysis and simulation to demonstrate the existence of otherwise undiscovered performance issues. We also demonstrate that by collecting dynamic phase profiles, we are able to approximate the analysis results of trace analysis and simulation, and more accurately represent the performance behavior of complex parallel applications than through flat or callpath profiles alone.

Modern parallel performance measurement systems collect performance information either through probes inserted in the application code or via statistical sampling. Probe-based techniques measure performance metrics directly using calls to a measurement library that execute as part of the application. In contrast, sampling-based systems interrupt program execution to sample metrics for statistical analysis of performance. Although both measurement approaches are represented by robust tool frameworks in the performance community, each has its strengths and weaknesses. In this paper, we investigate the creation of a hybrid measurement system, the goal being to exploit the strengths of both systems and mitigate their weaknesses. We show how such a system can be used to provide the application programmer with a more complete analysis of their application. Simple example and application codes are used to demonstrate its capabilities. We also show how the hybrid techniques can be combined to provide real cross-language performance evaluation of an uninstrumented run for mixed compiled/interpreted execution environments (e.g., Python and C/C++/Fortran).

Automating the process of parallel performance experimentation, analysis, and problem diagnosis can enhance environments for performance-directed application development, compilation, and execution. This is especially true when parametric studies, modeling, and optimization strategies require large amounts of data to be collected and processed for knowledge synthesis and reuse. This paper describes the integration of the PerfExplorer performance data mining framework with the OpenUH compiler infrastructure. OpenUH provides auto-instrumentation of source code for performance experimentation and PerfExplorer provides automated and reusable analysis of the performance data through a scripting interface. More importantly, PerfExplorer inference rules have been developed to recognize and diagnose performance characteristics important for optimization strategies and modeling. Three case studies are presented which show our success with automation in OpenMP and MPI code tuning, parametric characterization, Pand power modeling. The paper discusses how the integration supports performance knowledge engineering across applications and feedback-based compiler optimization in general.

Empirical performance evaluation of parallel systems and applications can generate significant amounts of performance data and analysis results from multiple experiments as performance is investigated and problems diagnosed. Hence, the management of performance information is a core component of performance analysis tools. To better support tool integration, portability; and reuse, there is a strong motivation to develop performance data management technology that can provide a common foundation for performance data storage, access, merging, and analysis. This paper presents the design and implementation of the performance data management framework (PerfDMF). PerfDMF addresses objectives of performance tool integration, interoperation, and reuse by providing common data storage, access, and analysis infrastructure for parallel performance profiles. PerfDMF includes an extensible parallel profile data schema and relational database schema, a profile query and analysis programming interface, and an extendible toolkit for profile import/export and standard analysis. We describe the PerfDMF objectives and architecture, give detailed explanation of the major components, and show examples of PerfDMF application.

Parallel applications running on high-end computer systems manifest a complexity of performance phenomena. Tools to observe parallel performance attempt to capture these phenomena in measurement datasets rich with information relating multiple performance metrics to execution dynamics and parameters specific to the application-system experiment. However, the potential size of datasets and the need to assimilate results from multiple experiments makes it a daunting challenge to not only process the information, but discover and understand performance insights. In this paper, we present PerfExplorer, a framework for parallel performance data mining and knowledge discovery. The framework architecture enables the development and integration of data mining operations that will be applied to large-scale parallel performance profiles. PerfExplorer operates as a client-server system and is built on a robust parallel performance database (PerfDMF) to access the parallel profiles and save its analysis results. Examples are given demonstrating these techniques for performance analysis of ASCI applications.

PerfTrack is a data store and interface for managing performance data from large-scale parallel applications. Data collected in different locations and formats can be compared and viewed in a single performance analysis session. The underlying data store used in PerfTrack is implemented with a database management system (DBMS). PerfTrack includes interfaces to the data store and scripts for automatically collecting data describing each experiment, such as build and platform details. We have implemented a prototype of PerfTrack that can use Oracle or PostgreSQL for the data store. We demonstrate the prototype's functionality with three case studies: one is a comparative study of an ASC purple benchmark on high-end Linux and AIX platforms; the second is a parameter study conducted at Lawrence Livermore National Laboratory (LLNL) on two high end platforms, a 128 node cluster of IBM Power 4 processors and BlueGene/L; the third demonstrates incorporating performance data from the Paradyn Parallel Performance Tool into an existing PerfTrack data store.