Statistical Performance Tuning of Parallel Monte Carlo Ocean Color Simulations

Kajiyama, Tamito; D’Alimonte, Davide; Cunha, José C.

doi:10.1109/pdcat.2012.125

Cited by 4 publications

(8 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another example comes from High Performance Computing, where it is common to change an application parameter to adapt a running application. In [40] a threshold value is changed while executing parallel Monte Carlo ocean color simulations, while [18] presents a study on tuning Fast Fourier Transformations on graphic processing units. Also, Rahman et al [56] and Tiwari et al [65] studied the effect of compiler parameters on both performance and power/energy consumption for scientific computing.…”

Section: Related Workmentioning

confidence: 99%

Automated design of self-adaptive software with control-theoretical formal guarantees

Filieri

Hoffmann

Maggio

2014

Proceedings of the 36th International Conference on Software Engineering

121

123

View full text Add to dashboard Cite

Self-adaptation enables software to execute successfully in dynamic, unpredictable, and uncertain environments.Control theory provides a broad set of mathematically grounded techniques for adapting the behavior of dynamic systems. While it has been applied to specific software control problems, it has proved difficult to define methodologies allowing non-experts to systematically apply control techniques to create adaptive software. These difficulties arise because computer systems are usually non-linear, with varying workloads and heterogeneous components, making it difficult to model software as a dynamic system; i.e., by means of differential or difference equations. This paper proposes a broad scope methodology for automatically constructing both an approximate dynamic model of a software system and a suitable controller for managing its non-functional requirements. Despite its generality, this methodology provides formal guarantees concerning the system's dynamic behavior by keeping its model continuously updated to compensate for changes in the execution environment and effects of the initial approximation.We apply the methodology to three case studies, demonstrating its generality by tackling different domains (and different non-functional requirements) with the same approach. Being broadly applicable and fully automated, this methodology may allow the adoption of control theoretical solutions (and their formal properties) for a wide range of software adaptation problems.

show abstract

Section: Related Workmentioning

confidence: 99%

Automated design of self-adaptive software with control-theoretical formal guarantees

Filieri

Hoffmann

Maggio

2014

Proceedings of the 36th International Conference on Software Engineering

121

123

View full text Add to dashboard Cite

show abstract

“…By the same token, the Mie scattering, included for a more realistic representation of light polarization effects, provides patterns analogous to published results [40]. As in former investigations [14][15][16], high-performance computing methods have been strategic to satisfy Monte Carlo simulation requirements by relying on a large-scale computer cluster for production runs. The Navigator supercomputer, University of Coimbra, Portugal, has been utilized in the present study.…”

Section: Discussionmentioning

confidence: 53%

“…Specific ρ variations due to the light polarization and changes in the sea-surface statistics are then detailed. This work relies on the MC code for Ocean Color Simulations MOX to model the reflectance factor in a three-dimensional domain using state-of-the-art High Performance Computing HPC solutions [14][15][16]. Above-water simulation results integrate former MOX investigations on the precision of data products derived with in-water radiometric systems [17,18].…”

Section: Introductionmentioning

confidence: 99%

Effects of light polarization and waves slope statistics on the reflectance factor of the sea surface

D’Alimonte¹,

Kajiyama²

2016

Opt. Express

Self Cite

View full text Add to dashboard Cite

Above-water radiometry depends on estimates of the reflectance factor ρ of the sea surface to compute the in situ water-leaving radiance. The Monte Carlo code for ocean color simulations MOX is used in this study to analyze the effect of different environmental components on ρ values. A first aspect is examining the reflectance factor without and by accounting for the sky-radiance polarization. The influence of the sea-surface statistics at discrete grid points is then considered by presenting a new scheme to define the variance of the waves slope. Results at different sun elevations and sensor orientations indicate that the light polarization effect on ρ simulations reduces from ∼17 to ∼10% when the wind speed increases from 0 to 14m s^-1. An opposite tendency characterizes the modeling of the sea-surface slope variance, with ρ differences up to ∼12% at a wind speed of 10m s^-1. The joint effect of polarization and the the sea-surface statistics displays a less systematic dependence on the wind speed, with differences in the range ∼13 to ∼18%. The ρ changes due to the light polarization and the variance of the waves slope become more relevant at sky-viewing geometries respectively lower and higher than 40° with respect to the zenith. An overall compensation of positive and negative offsets due to light polarization is finally documented when considering different sun elevations. These results address additional investigations which, by combining the modeling and experimental components of marine optics, better evaluate specific measurement protocols for collecting above-water radiometric data in the field.

show abstract

“…The present work advances reference research achievements by integrating new results and methods in a unifying framework of performance optimization techniques. Numerical results acknowledge an extended range of input simulation parameters to investigate the effects of environmental factors on run‐time performance optimization.…”

Section: Introductionmentioning

confidence: 97%

“…Within the scope of the Geo‐Info project aiming to promote joint research between computer scientists and Geoscience experts, this paper presents an HPC framework for performance prediction and optimization of Monte Carlo (MC) simulations for ocean color (OC) applications. This framework concerns a parallel MC radiative transfer code , hereafter referred to as MOX, developed by the authors to support investigations on OC data products derived from in situ radiometric measurements perturbed by uncertainties due to environmental factors . The MOX implementation is underpinned by the increasing importance of accurate long‐term in situ OC measurements for the development of bio‐optical models, vicarious calibration of satellite sensors, and validation of remote sensing (RS) radiometric products .…”

Section: Introductionmentioning

confidence: 99%

A high‐performance computing framework for Monte Carlo ocean color simulations

Kajiyama

D’Alimonte

Cunha

2016

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Summary This paper presents a high‐performance computing (HPC) framework for Monte Carlo (MC) simulations in the ocean color (OC) application domain. The objective is to optimize a parallel MC radiative transfer code named MOX, developed by the authors to create a virtual marine environment for investigating the quality of OC data products derived from in situ measurements of in‐water radiometric quantities. A consolidated set of solutions for performance modeling, prediction, and optimization is implemented to enhance the efficiency of MC OC simulations on HPC run‐time infrastructures. HPC, machine learning, and adaptive computing techniques are applied taking into account a clear separation and systematic treatment of accuracy and precision requirements for large‐scale MC OC simulations. The added value of the work is the integration of computational methods and tools for MC OC simulations in the form of an HPC‐oriented problem‐solving environment specifically tailored to investigate data acquisition and reduction methods for OC field measurements. Study results highlight the benefit of close collaboration between HPC and application domain researchers to improve the efficiency and flexibility of computer simulations in the marine optics application domain. © 2016 The Authors. Concurrency and Computation: Practice and Experience Published by John Wiley & Sons, Ltd.

show abstract

Statistical Performance Tuning of Parallel Monte Carlo Ocean Color Simulations

Cited by 4 publications

References 14 publications

Automated design of self-adaptive software with control-theoretical formal guarantees

Automated design of self-adaptive software with control-theoretical formal guarantees

Effects of light polarization and waves slope statistics on the reflectance factor of the sea surface

A high‐performance computing framework for Monte Carlo ocean color simulations

Contact Info

Product

Resources

About