Search NASA⌕ Search

SEARCH · Search NASA

Results for “performance autotuning”

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

Parallelizing autotuning for HPC applications: Unveiling the potential of the speculation strategy in Bayesian optimization

In the exascale computing era, tuning High-Performance Computing (HPC) applications has become a significant computational challenge. Although Bayesian optimization (BO) has emerged as a promising tool for HPC performance tuning, the BO workflow is inherently sequential (i.e., one function evaluation at a time) and cannot leverage the huge amount of parallel resources present in modern supercomputers, resulting in a considerable underutilization of their computational capabilities. This paper explores the trade-off between search quality and parallelism in BO, investigating a diverse set of methods. Building upon both previous approaches from the literature and novel methodologies introduced in this work, our study provides a deep analysis to accelerate BO performance tuning. By examining a set of synthetic functions and practical HPC applications, our exploration analyzes the interaction among various BO methods for parallelization, the quantity of parallel resources, the runtime distribution of target HPC applications, and the costs associated with different search orchestration mechanisms that have been overlooked in previous studies. Compared to sequential BO, our novel methodology achieves comparable quality while demonstrating robust scalability in search time as the amount of parallel resources increases; it also outperforms a state-of-the-art tuner, which supports parallelization, achieving up to 3.67x faster search time. We provide high-value insights for practitioners seeking to leverage the power of parallel computing for efficient HPC application tuning. Additionally, to further assist researchers in accelerating the performance tuning of their HPC applications, we provide an extension of an existing open-source tuning framework that incorporates our methods.

Bayesian optimization↗

Experiences in autotuning matrix multiplication for energy minimization on GPUs

In this study, we report extensive results and analysis of autotuning the computationally intensive graphics processing units kernel for dense matrix–matrix multiplication in double precision. In contrast to traditional autotuning and/or optimization for runtime performance only, we also take the energy efficiency into account. For kernels achieving equal performance, we show significant differences in their energy balance. We also identify the memory throughput as the most influential metric that trades off performance and energy efficiency. Finally, as a result, the performance optimal case ends up not being the most efficient kernel in overall resource use.

97 MATHEMATICS AND COMPUTING↗

Compiler-based code generation and autotuning for geometric multigrid on GPU-accelerated supercomputers

GPUs, with their high bandwidths and computational capabilities are an increasingly popular target for scientific computing. Unfortunately, to date, harnessing the power of the GPU has required use of a GPU-specific programming model like CUDA, OpenCL, or OpenACC. Thus, in order to deliver portability across CPU-based and GPU-accelerated supercomputers, programmers are forced to write and maintain two versions of their applications or frameworks. In this paper, we explore the use of a compiler-based autotuning framework based on CUDA-CHiLL to deliver not only portability, but also performance portability across CPU- and GPU-accelerated platforms for the geometric multigrid linear solvers found in many scientific applications. We also show that with autotuning we can attain near Roofline (a performance bound for a computation and target architecture) performance across the key operations in the miniGMG benchmark for both CPU- and GPU-based architectures as well as for a multiple stencil discretizations and smoothers. We show that our technology is readily interoperable with MPI resulting in performance at scale equal to that obtained via hand-optimized MPI+CUDA implementation.

97 MATHEMATICS AND COMPUTING↗

Simulating the weak death of the neutron in a femtoscale universe with near-Exascale computing

The fundamental particle theory called Quantum Chromodynamics (QCD) dictates everything about protons and neutrons, from their intrinsic properties to interactions that bind them into atomic nuclei. Quantities that cannot be fully resolved through experiment, such as the neutron lifetime (whose precise value is important for the existence of light-atomic elements that make the sun shine and life possible), may be understood through numerical solutions to QCD. We directly solve QCD using Lattice Gauge Theory and calculate nuclear observables such as neutron lifetime. We have developed an improved algorithm that exponentially decreases the time-tosolution and applied it on the new CORAL supercomputers, Sierra and Summit. We use run-time autotuning to distribute GPU resources, achieving 20% performance at low node count. We also developed optimal application mapping through a job manager, which allows CPU and GPU jobs to be interleaved, yielding 15% of peak performance when deployed across large fractions of CORAL.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Integrating ytopt and libEnsemble to autotune OpenMC

Ytopt is a Python machine-learning-based autotuning software package developed within the ECP PROTEAS-TUNE project. The ytopt software adopts an asynchronous search framework that consists of sampling a small number of input parameter configurations and progressively fitting a surrogate model over the input-output space until exhausting the user-defined maximum number of evaluations or the wall-clock time. libEnsemble is a Python toolkit for coordinating workflows of asynchronous and dynamic ensembles of calculations across massively parallel resources developed within the ECP PETSc/TAO project. libEnsemble helps users take advantage of massively parallel resources to solve design, decision, and inference problems and expands the class of problems that can benefit from increased parallelism. In this paper we present our methodology and framework to integrate ytopt and libEnsemble to take advantage of massively parallel resources to accelerate the autotuning process. Specifically, we focus on using the proposed framework to autotune the ECP ExaSMR application OpenMC, an open source Monte Carlo particle transport code. OpenMC has seven tunable parameters some of which have large ranges such as the number of particles in-flight, which is in the range of 100,000 to 8 million, with its default setting of 1 million. Setting the proper combination of these parameter values to achieve the best performance is extremely time-consuming. Therefore, we apply the proposed framework to autotune the MPI/OpenMP offload version of OpenMC based on a user-defined metric such as the figure of merit (FoM) (particles/s) or energy efficiency energy-delay product (EDP) on Crusher at Oak Ridge Leadership Computing Facility. In conclusion, the experimental results show that we achieve the improvement up to 29.49% in FoM and up to 30.44% in EDP.

Autotuning↗

Surrogate-Based Autotuning for Randomized Sketching Algorithms in Regression Problems

Algorithms from Randomized Numerical Linear Algebra (RandNLA) are known to be effective in handling high-dimensional computational problems, providing high-quality empirical performance as well as strong probabilistic guarantees. However, their practical application is complicated by the fact that the user needs to set various algorithm-specific tuning parameters which are different from those used in traditional NLA. This paper demonstrates how a surrogate-based autotuning approach can be used to address fundamental problems of parameter selection in RandNLA algorithms. In particular, we provide a detailed investigation of surrogate-based autotuning for sketch-and-precondition (SAP)-based randomized least squares methods, which have been one of the great success stories in modern RandNLA. Empirical results show that our surrogate-based autotuning approach can achieve near-optimal performance with much less tuning cost than a random search (up to about 7.6x fewer trials of different parameter configurations). Moreover, while our experiments focus on least squares, our results demonstrate a general-purpose autotuning pipeline applicable to any kind of RandNLA algorithm.

Cho, Younghyun↗

A multi-backend autotuning study of feature selection on GPUs

Abstract Feature selection is an important step in machine learning that can benefit from GPU acceleration. As the number of GPU vendors increases, it is imperative to adapt algorithms such as the minimum Redundancy Maximum Relevance (mRMR) feature selection method to different backends that support several GPU architectures. This work presents a multi-backend implementation of mRMR across CUDA, HIP, and SYCL, and studies its performance when combined with Bayesian optimization and transfer learning to automatically tune execution parameters for different platforms and datasets. Our experimental results show that when tuned, CUDA and HIP achieve comparable performance on NVIDIA architectures, while SYCL exhibits a moderate performance gap. Overall, this work highlights the impact of backend choice and autotuning on GPU-accelerated feature selection and provides insights into deploying mRMR across heterogeneous environments.

Beceiro, Bieito (ORCID:0000000333014890)↗

Collective I/O Tuning Using Analytical and Machine-Learning Models

The ever larger demand of scientific applications for computation and data is currently driving a continuous increase in scale of parallel computers. The inherent complexity of scaling up a computing systems in terms of both hardware and software stack exposes an increasing number of factors impacting the performance and complicating the process of optimization. In particular, the optimization of parallel I/O has become increasingly challenging due to increasing storage hierarchy and well known performance variability of shared storage systems. This paper focuses on model-based autotuning of the two-phase collective I/O algorithm from a popular MPI distribution on the Blue Gene/Q architecture. We propose a novel hybrid model, constructed as a composition of analytical models for communication and storage operations and black-box models for the performance of the individual operations. We perform an in-depth study of the complexity involved in performance modeling including architecture, software stack and noise. In particular we address this challenges of modeling the performance of shared storage systems by building a benchmark that helps synthesizing factors such as topology, file caching, and noise. The experimental results show that the hybrid approach produces significantly better results than state-of-the-art machine learning approaches and shows a higher robustness to noise, at the cost of a higher modeling complexity

I/O performance modeling↗

Critical Points Based Register-Concurrency Autotuning for GPUs

The unprecedented prevalence of GPGPU is largely attributed to its abundant on-chip register resources, which allow massively concurrent threads and extremely fast context switch. However, due to internal memory capacity constraints, there is a tradeoff between the per-thread register usage and the overall concurrency. This becomes a design problem in terms of performance tuning, since the performance “sweet spot” which can be significantly affected by these two factors is generally unknown beforehand. In this paper, we propose an effective autotuning solution to quickly and efficiently select the optimal number of registers perthread for delivering the best GPU performance. Experiments on three generations of GPUs (Nvidia Fermi, Kepler and Maxwell) demonstrate that our simple strategy can achieve an average of 10% performance improvement while a max of 50% over the original version without modifying the user program. Additionally, to reduce local cache misses due to register spilling and further improve performance, we explore three optimization schemes (i.e. bypass L1 for global memory access, enlarge local L1 cache and spill into shared memory) and discuss their impact on performance on a Kepler GPU.

auto-tuning↗

Performance Engineering Research Institute SciDAC-2 Enabling Technologies Institute (Final Report)

Enhancing the performance of SciDAC applications on petascale systems had high priority within DOE SC at the start of the second phase of the SciDAC program, SciDAC-2, as it continues to do so today. Achieving expected levels of performance on high-end computing (HEC) systems is growing ever more challenging due to enormous scale, increasing architectural complexity, and increasing application complexity. To address these challenges, the University of Southern California's Information Sciences Institute organized the Performance Engineering Research Institute (PERI). PERI implemented a unified, tripartite research plan encompassing: (1) performance modeling and prediction; (2) automatic performance tuning; and (3) performance engineering of high profile applications. Within PERI, USC's primary research activity was automatic tuning (autotuning) of scientific software. This activity was spurred by the strong user preference for automatic tools and was based on previous successful activities such as ATLAS, which automatically tuned components of the LAPACK linear algebra library, and other recent work on autotuning domain-specific libraries. Our other major component was application engagement, to which we devoted approximately 30% of our effort to work directly with SciDAC-2 applications. This report is a summary of the overall results of the USC PERI effort.

97 MATHEMATICS AND COMPUTING↗

Final Report for Award #DE-SC3956 Separating Algorithm and Implementation via programming Model Injection (SAIMI)

Programming parallel machines is fraught with difficulties: the obfuscation of algorithms due to implementation details such as communication and synchronization, the need for transparency between language constructs and performance, the difficulty of performing program analysis to enable automatic parallelization techniques, and the existence of important "dusty deck" codes. The SAIMI project developed abstractions that enable the orthogonal specification of algorithms and implementation details within the context of existing DOE applications. The main idea is to enable the injection of small programming models such as expressions involving transcendental functions, polyhedral iteration spaces with sparse constraints, and task graphs into full programs through the use of pragmas. These smaller, more restricted programming models enable orthogonal specification of many implementation details such as how to map the computation on to parallel processors, how to schedule the computation, and how to allocation storage for the computation. At the same time, these small programming models enable the expression of the most computationally intense and communication heavy portions in many scientific simulations. The ability to orthogonally manipulate the implementation for such computations will significantly ease performance programming efforts and expose transformation possibilities and parameter to automated approaches such as autotuning. At Colorado State University, the SAIMI project was supported through DOE grant DE-SC3956 from April 2010 through August 2015. The SAIMI project has contributed a number of important results to programming abstractions that enable the orthogonal specification of implementation details in scientific codes. This final report summarizes the research that was funded by the SAIMI project.

97 MATHEMATICS AND COMPUTING↗

GPTune v.1

GPTune is an autotuning framework for exascale applications. Given an application with a set of parameters controlling its performance (e.g. runtime, memory and/or energy consumption, numerical accuracy, ...), GPTune aims at finding a combination of the parameters that yields the best possible performance. Although general enough, GPTune specializes in the tunning of expensive-to-run applications, where the cost of running the application is very expensive (in terms of computing power requirement and/or runtime).

Demmel, JamesW.↗

ZettaBricks: A Language Compiler and Runtime System for Anyscale Computing

This grant supported the ZettaBricks and OpenTuner projects. ZettaBricks is a new implicitly parallel language and compiler where defining multiple implementations of multiple algorithms to solve a problem is the natural way of programming. ZettaBricks makes algorithmic choice a first class construct of the language. Choices are provided in a way that also allows our compiler to tune at a finer granularity. The ZettaBricks compiler autotunes programs by making both fine-grained as well as algorithmic choices. Choices also include different automatic parallelization techniques, data distributions, algorithmic parameters, transformations, and blocking. Additionally, ZettaBricks introduces novel techniques to autotune algorithms for different convergence criteria. When choosing between various direct and iterative methods, the ZettaBricks compiler is able to tune a program in such a way that delivers near-optimal efficiency for any desired level of accuracy. The compiler has the flexibility of utilizing different convergence criteria for the various components within a single algorithm, providing the user with accuracy choice alongside algorithmic choice. OpenTuner is a generalization of the experience gained in building an autotuner for ZettaBricks. OpenTuner is a new open source framework for building domain-specific multi-objective program autotuners. OpenTuner supports fully-customizable configuration representations, an extensible technique representation to allow for domain-specific techniques, and an easy to use interface for communicating with the program to be autotuned. A key capability inside OpenTuner is the use of ensembles of disparate search techniques simultaneously; techniques that perform well will dynamically be allocated a larger proportion of tests.

97 MATHEMATICS AND COMPUTING↗

Autotune E+ Building Energy Models

This paper introduces a novel Autotune methodology under development for calibrating building energy models (BEM). It is aimed at developing an automated BEM tuning methodology that enables models to reproduce measured data such as utility bills, sub-meter, and/or sensor data accurately and robustly by selecting best-match E+ input parameters in a systematic, automated, and repeatable fashion. The approach is applicable to a building retrofit scenario and aims to quantify the trade-offs between tuning accuracy and the minimal amount of ground truth data required to calibrate the model. Autotune will use a suite of machine-learning algorithms developed and run on supercomputers to generate calibration functions. Specifically, the project will begin with a de-tuned model and then perform Monte Carlo simulations on the model by perturbing the uncertain parameters within permitted ranges. Machine learning algorithms will then extract minimal perturbation combinations that result in modeled results that most closely track sensor data. A large database of parametric EnergyPlus (E+) simulations has been made publicly available. Autotune is currently being applied to a heavily instrumented residential building as well as three light commercial buildings in which a de-tuned model is autotuned using faux sensor data from the corresponding target E+ model.

automatic calibration↗

Final Report on XStack: Software Synthesis for High Productivity ExaScale Computing

The goal of the project was to develop a programming model that would significantly improve productivity in the high-performance computing domain by bringing together three components: a) Automated equivalence checking, b) Sketch-based program synthesis, and c) Autotuning. The report provides an executive summary of the research accomplished through this project. At the end of the report is appended a paper that describes in more detail the key technical accomplishments from this project, and which was published in SC 2014.

97 MATHEMATICS AND COMPUTING↗

Hydrogen maser frequency standard computer model for automatic cavity tuning servo simulations

A computer model of the JPL hydrogen maser frequency standard was developed. This model allows frequency stability data to be generated, as a function of various maser parameters, many orders of magnitude faster than these data can be obtained by experimental test. In particular, the maser performance as a function of the various automatic tuning servo parameters may be readily determined. Areas of discussion include noise sources, first-order autotuner loop, second-order autotuner loop, and a comparison of the loops.

Potter, P. D.↗

Oak Ridge Institutional Cluster Autotune Test Drive Report

The Oak Ridge Institutional Cluster (OIC) provides general purpose computational resources for the ORNL staff to run computation heavy jobs that are larger than desktop applications but do not quite require the scale and power of the Oak Ridge Leadership Computing Facility (OLCF). This report details the efforts made and conclusions derived in performing a short test drive of the cluster resources on Phase 5 of the OIC. EnergyPlus was used in the analysis as a candidate user program and the overall software environment was evaluated against anticipated challenges experienced with resources such as the shared memory-Nautilus (JICS) and Titan (OLCF). The OIC performed within reason and was found to be acceptable in the context of running EnergyPlus simulations. The number of cores per node and the availability of scratch space per node allow non-traditional desktop focused applications to leverage parallel ensemble execution. Although only individual runs of EnergyPlus were executed, the software environment on the OIC appeared suitable to run ensemble simulations with some modifications to the Autotune workflow. From a standpoint of general usability, the system supports common Linux libraries, compilers, standard job scheduling software (Torque/Moab), and the OpenMPI library (the only MPI library) for MPI communications. The file system is a Panasas file system which literature indicates to be an efficient file system.

99 GENERAL AND MISCELLANEOUS↗

StreamJIT: a commensal compiler for high-performance stream programming

There are many domain libraries, but despite the performance benefits of compilation, domain-specific languages are comparatively rare due to the high cost of implementing an optimizing compiler. We propose commensal compilation, a new strategy for compiling embedded domain-specific languages by reusing the massive investment in modern language virtual machine platforms. Commensal compilers use the host language's front-end, use host platform APIs that enable back-end optimizations by the host platform JIT, and use an autotuner for optimization selection. The cost of implementing a commensal compiler is only the cost of implementing the domain-specific optimizations. We demonstrate the concept by implementing a commensal compiler for the stream programming language StreamJIT atop the Java platform. Our compiler achieves performance 2.8 times better than the StreamIt native code (via GCC) compiler with considerably less implementation effort.

97 MATHEMATICS AND COMPUTING↗