Search NASA⌕ Search

Engineering topics

Williams, S

Publications and source records attributed to Williams, S.

Instruction Roofline: An insightful visual performance model for GPUs

The Roofline performance model provides an intuitive approach to identify performance bottlenecks and guide performance optimization. However, the classic FLOP-centric approach is inappropriate for the emerging applications that perform more integer operations than floating point operations. In this article, we reintroduce our Instruction Roofline Model on NVIDIA GPUs and expand our evaluation of it. The Instruction Roofline incorporates instructions and memory transactions across all memory hierarchies together, and provides more performance insights than the FLOP-oriented Roofline Model, that is, instruction throughput, stride memory access patterns, bank conflicts, and thread predication. We use our Instruction Roofline methodology to analyze eight proxy applications: HPGMG from AMReX, Matrix Transpose benchmarks, ADEPT from MetaHipMer's sequence alignment phase, EXTENSION from MetaHipMer's local assembly phase, CUSP, cuSPARSE, cudaTensorCoreGemm, and cuBLAS. We demonstrate the ability of our methodology to understand various aspects of performance and performance bottlenecks on NVIDIA GPUs and motivate code optimizations.

Ding, N↗

Roofline scaling trajectories: A method for parallel application and architectural performance analysis

The end of Dennard scaling signaled a shift in HPC supercomputer architectures from systems built from single-core processor architectures to systems built from multicore and eventually manycore architectures. This transition substantially complicated performance optimization and analysis as new programming models were created, new scaling methodologies deployed, and on-chip contention became a bottleneck to performance. Existing distributed memory performance models like logP and logGP were unable to capture this contention. The Roofline model was created to address this contention and its interplay with locality. However, to date, the Roofline model has focused on full-node concurrency. In this paper, we extend the Roofline model to capture the effects of concurrency on data locality and on-chip contention. We demonstrate the value of this new technique by evaluating the NAS parallel benchmarks on both multicore and manycore architectures under both strong-And weak-scaling regimes. In order to quantify the interplay between programming model and locality, we evaluate scaling under both the OpenMP and flat MPI programming models.

Ibrahim, K↗

Evaluating and optimizing the NERSC workload on Knights Landing

NERSC has partnered with 20 representative application teams to evaluate performance on the Xeon-Phi Knights Landing architecture and develop an application-optimization strategy for the greater NERSC workload on the recently installed Cori system. In this article, we present early case studies and summarized results from a subset of the 20 applications highlighting the impact of important architecture differences between the Xeon-Phi and traditional Xeon processors. We summarize the status of the applications and describe the greater optimization strategy that has formed.

97 MATHEMATICS AND COMPUTING↗

Evaluation of PGAS communication paradigms with geometric multigrid

Partitioned Global Address Space (PGAS) languages and one-sided communication enable application developers to select the communication paradigm that balances the performance needs of applications with the productivity desires of programmers. In this paper, we evaluate three different one-sided communication paradigms in the context of geometric multigrid using the miniGMG benchmark. Although miniGMG's static, regular,and predictable communication does not exploit the ultimate potential of PGAS models, multigrid solvers appear in many contemporary applications and represent one of the most important communication patterns. We use UPC++, a PGAS extension of C++, as the vehicle for our evaluation, though our work is applicable to any of the existing PGAS languages and models. We compare performance with the highly tuned MPI baseline, and the results indicate that the most promising approach towards achieving performance and ease of programming is to use high-level abstractions, such as the multidimensional arrays provided by UPC++, that hide data aggregation and messaging in the runtime library.

Shan, H↗

High-Efficiency Retrofit Lessons for Retail from a SuperTarget: Preprint

The National Renewable Energy Laboratory partnered with Target under the Commercial Building Program to design and implement a retrofit of a SuperTarget in Thornton, CO. The result was a retrofit design that predicted 37% energy savings over ASHRAE Standard 90.1-2004, and 29% compared to existing (pre-retrofit) store consumption. The largest savings came from energy efficient lighting, energy efficient cooling systems, improved refrigeration, and better control of plug loads.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

The VAMOS Ocean-Cloud-Atmosphere-Land Study Regional Experiment (VOCALS-REx): Goals, platforms, and field operations

The VAMOS Ocean-Cloud-Atmosphere-Land Study Regional Experiment (VOCALS-REx) was an international field program designed to make observations of poorly understood but critical components of the coupled climate system of the southeast Pacific. This region is characterized by strong coastal upwelling, the coolest SSTs in the tropical belt, and is home to the largest subtropical stratocumulus deck on Earth. The field intensive phase of VOCALS-REx took place during October and November 2008 and constitutes a critical part of a broader CLIVAR program (VOCALS) designed to develop and promote scientific activities leading to improved understanding, model simulations, and predictions of the southeastern Pacific (SEP) coupled ocean-atmosphere-land system, on diurnal to interannual timescales. The other major components of VOCALS are a modeling program with a model hierarchy ranging from the local to global scales, and a suite of extended observations from regular research cruises, instrumented moorings, and satellites. The two central themes of VOCALS-REx focus upon (a) links between aerosols, clouds and precipitation and their impacts on marine stratocumulus radiative properties, and (b) physical and chemical couplings between the upper ocean and the lower atmosphere, including the role that mesoscale ocean eddies play. A set of hypotheses designed to be tested with the combined field, monitoring and modeling work in VOCALS is presented here. A further goal of VOCALS-REx is to provide datasets for the evaluation and improvement of large-scale numerical models. VOCALS-REx involved five research aircraft, two ships and two surface sites in northern Chile. We describe the instrument payloads and key mission strategies for these platforms and give a summary of the missions conducted.

54 ENVIRONMENTAL SCIENCES↗

Methane ignition catalyzed by in situ generated palladium nanoparticles

Catalytic ignition of methane over the surfaces of freely-suspended and in situ generated palladium nanoparticles was investigated experimentally and numerically. The experiments were conducted in a laminar flow reactor. The palladium precursor was a compound (Pd(THD){sub 2}, THD: 2,2,6,6-tetramethyl-3,5-heptanedione) dissolved in toluene and injected into the flow reactor as a fine aerosol, along with a methane-oxygen-nitrogen mixture. For experimental conditions chosen in this study, non-catalytic, homogeneous ignition was observed at a furnace temperature of {proportional_to}1123 K, whereas ignition of the same mixture with the precursor was found to be {proportional_to}973 K. In situ production of Pd/PdO nanoparticles was confirmed by scanning mobility, transmission electron microscopy and X-ray photoelectron spectroscopy analyses of particles collected at the reactor exit. The catalyst particle size distribution was log-normal. Depending on the precursor loading, the median diameter ranged from 10 to 30 nm. The mechanism behind catalytic ignition was examined using a combined gas-phase and gas-surface reaction model. Simulation results match the experiments closely and suggest that palladium nanocatalyst significantly shortens the ignition delay times of methane-air mixtures over a wide range of conditions. (author)

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

PERI auto-tuning.

The enormous and growing complexity of today's high-end systems has increased the already significant challenges of obtaining high performance on equally complex scientific applications. Application scientists are faced with a daunting challenge in tuning their codes to exploit performance-enhancing architectural features. The Performance Engineering Research Institute (PERI) is working toward the goal of automating portions of the performance tuning process. This paper describes PERI's overall strategy for auto-tuning tools and recent progress in both building auto-tuning tools and demonstrating their success on kernels, some taken from large-scale applications.

99 GENERAL AND MISCELLANEOUS↗