Performance Modeling, Benchmarking and Simulation of High Performance Computing Systems

Wright, Steven A.

doi:10.1016/j.future.2018.11.020

Cited by 17 publications

(10 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Parallel computing is being widely studied and has been implemented using different approaches in recent years and it is expected that the threshold of ExaFLOP (1018 floating-point operations per second) will be reached by the year 2020. Thus, parallel computers could replace the current computers, which mostly have a sequential architecture (Li, 2018;Wright, 2019). However, silicon and molecular computers continue to use binary logic, and this force translating non-binary operations into binary atomic operations using Boolean algebra (Zhang et al, 2019;Eshra et al, 2019).…”

Section: Discussionmentioning

confidence: 99%

A new DNA-based model for finite field arithmetic

Jirón

Soto

Marín

et al. 2019

Heliyon

View full text Add to dashboard Cite

A B S T R A C TA Galois field GFðp n Þ with p ! 2 a prime number and n ! 1 is a mathematical structure widely used in Cryptography and Error Correcting Codes Theory. In this paper, we propose a novel DNA-based model for arithmetic over GFðp n Þ. Our model has three main advantages over other previously described models. First, it has a flexible implementation in the laboratory that allows the realization arithmetic calculations in parallel for p ! 2, while the tile assembly and the sticker models are limited to p ¼ 2. Second, the proposed model is less prone to error, because it is grounded on conventional Polymerase Chain Reaction (PCR) amplification and gel electrophoresis techniques. Hence, the problems associated to models such as tile-assembly and stickers, that arise when using more complex molecular techniques, such as hybridization and denaturation, are avoided. Third, it is simple to implement and requires 50 ng/μL per DNA double fragment used to develop the calculations, since the only feature of interest is the size of the DNA double strand fragments. The efficiency of our model has execution times of order Oð1Þ and OðnÞ, for the addition and multiplication over GFðp n Þ, respectively. Furthermore, this paper provides one of the few experimental evidences of arithmetic calculations for molecular computing and validates the technical applicability of the proposed model to perform arithmetic operations over GFðp n Þ.

show abstract

Section: Discussionmentioning

confidence: 99%

A new DNA-based model for finite field arithmetic

Jirón

Soto

Marín

et al. 2019

Heliyon

View full text Add to dashboard Cite

show abstract

“…This represents the fourth PMBS special issue, following two issues of the computer journal 13 and one issue of future generation computer systems. 14 The six accepted papers touch upon many of the areas of interest at the PMBS workshop, including modeling novel architectures, application runtime prediction, heuristic performance modeling and parallel performance optimization. The selected papers represent worldwide programmes of research focused on the performance of scientific applications on HPC systems.…”

Section: Concurrency and Computation: Practice And Experiencementioning

confidence: 99%

Special issue on Performance modeling, benchmarking, and simulation of high performance computing systems

Wright

2022

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

“…Computational expense is often raised as a barrier to operational numerical simulations. The state of the art in high‐performance computing is petascale (

>

^{15}

floating point operations per second), with exascale computing expected to be achieved in the near future (Wright, ). Innocenti et al () present an approach to space weather prediction that aims toward exascale computing capabilites, using particle‐in‐cell methods to resolve physics on kinetic scales everywhere in the magnetosphere.…”

Section: Challenges In Numerical Modelingmentioning

confidence: 99%

Challenges and Opportunities in Magnetospheric Space Weather Prediction

Morley

2020

Space Weather

View full text Add to dashboard Cite

Space Weather is the study of the dynamics of the coupled Solar‐Terrestrial environment, as these dynamics impact technological systems and human activity. This paper reviews a selection of the advances, challenges, and new opportunities for magnetospheric space weather, and while the focus is on specific phenomena, many other aspects of space weather will have similar challenges and needs. As a scientific field with direct applications, the field of space weather is partly driven by imperatives from both policy and operations. We provide an introduction to some of the context in which the field exists and discuss how this might shape future developments and norms within the space weather enterprise. We briefly examine benchmarking, as a policy and operationally driven activity, as it provides immediate societal relevance and an opportunity to stretch scientific understanding. As numerical space weather prediction now becomes routine, and exascale computing is in the near future, we identify challenges relating to computational expense and big data, capturing and accounting for uncertainties, and specification of boundary conditions. Here, as with the observations supporting numerical space weather prediction, the key challenge lies in extending the lead time of predictions. We also discuss the role of data, particularly in regard to model validation and empirical modeling. Due to the growing societal impact of space weather, we also examine the relationships between space weather and its terrestrial counterpart and look at the importance of continuous evaluation, monitoring progress in predictive capability, and communication with researchers, forecasters, and end users.

show abstract

Performance Modeling, Benchmarking and Simulation of High Performance Computing Systems

Cited by 17 publications

References 26 publications

A new DNA-based model for finite field arithmetic

A new DNA-based model for finite field arithmetic

Special issue on Performance modeling, benchmarking, and simulation of high performance computing systems

Challenges and Opportunities in Magnetospheric Space Weather Prediction

Contact Info

Product

Resources

About