Optimal Implementation of Two FIFO-Queues in Single-Level Memory

Aksenova, E. A.; Sokolov, A. V.

doi:10.4236/am.2011.210180

Cited by 7 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Preliminary results were reported in [19]. Previously, such models were built by our team for some other dynamic data structures: FIFO-queues, stacks, priority queues [9,10,15,17,18].…”

Section: Introductionmentioning

confidence: 99%

The Optimal Control of Two Work-Stealing Deques, Moving One After Another in a Shared Memory

Barkovsky¹,

Lazutina

Sokolov³

2019

ПСТП

View full text Add to dashboard Cite

In the parallel work-stealing load balancers, each core owns personal buffer of tasks called deque. One end of the deque is used by its owner to add and retrieve tasks, while the second end is used by other cores to steal tasks. In the paper two representation methods of deques are analyzed: partitioned serial cyclic representation of deques (one of the conventional techniques); and the new approach proposed by our team, without partition of shared memory in advance between deques moving one after another in a circle. Previously we analyzed these methods for representing FIFO queues in network applications, where the “One after another” way gave the best result for some values of the system parameters. Purpose of this research is to construct and analyze models of the process of work with two circular deques located in shared memory, where they movie one after another in a circle. The mathematical model is constructed in the form of a random walk by integer points in the pyramid. The simulation model is constructed using the Monte Carlo method. The used work-stealing strategy is stealing of one element. We propose the mathematical and simulation models of this process and carry out numerical experiments.

show abstract

“…Preliminary results were reported in [19]. Previously, such models were built by our team for some other dynamic data structures: FIFO-queues, stacks, priority queues [9,10,15,17,18].…”

Section: Introductionmentioning

confidence: 99%

The Optimal Control of Two Work-Stealing Deques, Moving One After Another in a Shared Memory

Barkovsky¹,

Lazutina

Sokolov³

2019

ПСТП

View full text Add to dashboard Cite

show abstract

“…In [5][6][7] models were offered for sequential, linked and paged ways of representation of multiple FIFO-queues in the memory of one level. These models assume that at each step of discrete time, with a given probabilities, some operations with data structures will occur.…”

Section: Introductionmentioning

confidence: 99%

Some problems of optimal control of two parallel FIFO-queues

Sokolov¹,

Barkovsky²

2015

AIP Conference Proceedings

Self Cite

View full text Add to dashboard Cite

This paper presents new mathematical models and algorithms of optimal control of multiple parallel FIFOqueues in the memory of one level, where the queues are moving consecutive and one after another in a circle. As mathematical models we propose random walks on the integer lattice in different areas on the plane and in 3-dimensional space. The optimality criterion is the minimum average portion of lost elements on infinite working time.

show abstract

“…Математическую модель мы будем строить как случайное блуждание по целым точкам в пирамиде, где переходы осуществляет процесс в дискретном времени с заданными вероятностями, соответствующими определенным операциям с деками. Раньше такие модели мы строили для других динамических структур данных: FIFO-очередей, стеков, приоритетных очередей [9,10,15,17,18]. Результаты предварительных исследований с деками докладывались в [19].…”

unclassified

The Optimal Control of Two Work-Stealing Deques, Moving One After Another in a Shared Memory

Барковский¹,

Lazutina²,

Sokolov³

2019

ПСТП

View full text Add to dashboard Cite

В work-stealing балансировщиках параллельных задач, каждое ядро имеет свой буфер задач—дек (англ. deque). Владелец дека использует один конец для добавления и извлечения задач, а из второго конца задачи перехватываются другими ядрами. В статье анализируются два метода представления деков: один из распространенных методов—раздельное последовательное циклическое представление деков; и новый предложенный нами метод, где общая память для деков заранее не делится и они двигаются друг за другом по кругу. Ранее эти методы анализировались нами для представления FIFO-очередей в сетевых приложениях, где для некоторых значений параметров системы метод «Друг за другом» давал лучший результат. Целью исследования является построение и анализ модели процесса работы с двумя последовательными деками, когда они двигаются друг за другом по кругу в общей памяти. Математическую модель мы будем строить как случайное блуждание по целым точкам в пирамиде. Имитационная модель строится с помощью метода Монте-Карло. Используемая стратегия work-stealing—перехват одного элемента. Предложены математическая и имитационная модели данного процесса и проведены численные эксперименты.

show abstract

Optimal Implementation of Two FIFO-Queues in Single-Level Memory

Cited by 7 publications

References 6 publications

The Optimal Control of Two Work-Stealing Deques, Moving One After Another in a Shared Memory

The Optimal Control of Two Work-Stealing Deques, Moving One After Another in a Shared Memory

Some problems of optimal control of two parallel FIFO-queues

The Optimal Control of Two Work-Stealing Deques, Moving One After Another in a Shared Memory

Contact Info

Product

Resources

About