Managing multi-configuration hardware via dynamic working set analysis

Dhodapkar, Ashutosh S.; Smith, James E.

doi:10.1145/545214.545241

Cited by 96 publications

(52 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…No contexto de consumo de recursos, diversos trabalhos estudam diferentes abordagens, desde métricas dinamicamente identificadas para otimização adaptativa [1,2,3], até técnicas que identificam cargas de trabalho através de simulação [9,23,22] Balasubramonian et al [1,2] explora mecanismos para reconfigurar dinamicamente uma micro-arquitetura de acordo com as necessidade da aplicação. A identificação de mudanças no comportamento do programa em tempo de execuçãoé feita com base nas instruções e acessoà memória.…”

Section: Trabalhos Relacionadosunclassified

“…A identificação de mudanças no comportamento do programa em tempo de execuçãoé feita com base nas instruções e acessoà memória. Dhodapkar e Smith [3] estendem o algoritmo utilizado em [1] para compor um conjunto de "assinaturas" de instruções de aplicações criando um mecanismo baseado em tabela, permitindo escolher a melhor configuração a ser utilizada no hardware reconfigurável a partir da caracterização da aplicação com base na comparação de assinaturas de execuções anteriores.…”

Section: Trabalhos Relacionadosunclassified

See 1 more Smart Citation

Caracterização do perfil de consumo de recursos de programas binários utilizando a técnica DAMICORE

Pinto

Delbem

Monaco

2017

Anais Do Simpósio Brasileiro De Sistemas De Informação (SBSI)

View full text Add to dashboard Cite

Resource consumption profile of a computer program is a relevant information with a wide scope of application such as schedulers, load balancers, Quality of Service (QoS) systems, among others. Conventional techniques available for this purpose include static source code analysis and profile matching based on runtime execution measurements. On the one hand, source code can't often be available for analysis, but on the other hand, make measurements on an execution system requires special precautions to minimize the overhead. This paper presents the use of DAMICORE data mining technique to promote the characterization of binary computer programs by their resource consumption profile. The technique was successfully applied over a dataset composed by 80 binary programs carefully selected.

show abstract

Section: Trabalhos Relacionadosunclassified

Caracterização do perfil de consumo de recursos de programas binários utilizando a técnica DAMICORE

Pinto

Delbem

Monaco

2017

Anais Do Simpósio Brasileiro De Sistemas De Informação (SBSI)

View full text Add to dashboard Cite

show abstract

“…Ponomarev, et al, uses prior occupancy to predict future occupancy of various queues every 2,000 cycles [27]. Dhodapkar, et al, assess the similarity of working sets to trigger reconfiguration [7]. Hughes et al consider adapting resources for multimedia workloads based on frame type while Huang, et al, consider adapting resources for code sections defined by subroutine boundaries [17,16].…”

Section: Related Workmentioning

confidence: 99%

“…The computational complexity of this problem has hindered advances in microarchitectural adaptivity as prior work often constrained adaptivity in the temporal (e.g., applications [24], working sets [7], subroutines [16], and multimedia frames [17]) and/or spatial (e.g., two or three parameters among depth [8,31], width [17], queues [1,27], and caches [24,4]) dimensions. Analyses with constraints to temporal adaptivity do not fully illustrate the potential efficiency gains of dynamic structural reconfiguration.…”

Section: Introductionmentioning

confidence: 99%

Efficiency trends and limits from comprehensive microarchitectural adaptivity

LeeBenjamin

BrooksDavid

2008

SIGOPS Oper. Syst. Rev.

View full text Add to dashboard Cite

Increasing demand for power-efficient, high-performance computing requires tuning applications and/or the underlying hardware to improve the mapping between workload heterogeneity and computational resources. To assess the potential benefits of hardware tuning, we propose a framework that leverages synergistic interactions between recent advances in (a) sampling, (b) predictive modeling, and (c) optimization heuristics. This framework enables qualitatively new capabilities in analyzing the performance and power characteristics of adaptive microarchitectures. For the first time, we are able to simultaneously consider high temporal and comprehensive spatial adaptivity. In particular, we optimize efficiency for many, short adaptive intervals and identify the best configuration of 15 parameters, which define a space of 240B points.With frequent sub-application reconfiguration and a fully reconfigurable hardware substrate, adaptive microarchitectures achieve bips 3 /w efficiency gains of up to 5.3x (median 2.4x) relative to their static counterparts already optimized for a given application. This 5.3x efficiency gain is derived from a 1.6x performance gain and 0.8x power reduction. Although several applications achieve a significant fraction of their potential efficiency with as few as three adaptive parameters, the three most significant parameters differ across applications. These differences motivate a hardware substrate capable of comprehensive adaptivity to meet these diverse application requirements.

show abstract

“…For example, Khan et al [57] used phase classification to facilitate thread-to-core assignment in asymmetric multicore systems, in order to optimize throughput, power, or performance per watt. Dhodapkar et al [32] used adaptable instruction caches to meet each phase's instruction execution needs for optimized performance efficiency.…”

Section: Introductionmentioning

confidence: 99%

A Survey of Phase Classification Techniques for Characterizing Variable Application Behavior

Criswell

Adegbija

2020

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

Adaptable computing is an increasingly important paradigm that specializes system resources to variable application requirements, environmental conditions, or user requirements. Adapting computing resources to variable application requirements (or application phases) is otherwise known as phase-based optimization. Phase-based optimization takes advantage of application phases, or execution intervals of an application that behave similarly, to enable effective and beneficial adaptability. In order for phase-based optimization to be effective, the phases must first be classified to determine when application phases begin and end, and ensure that system resources are accurately specialized. In this paper, we present a survey of phase classification techniques that have been proposed to exploit the advantages of adaptable computing through phase-based optimization. We focus on recent techniques, and classify these techniques with respect to several factors in order to highlight their similarities and differences. We divide the techniques by their major defining characteristics-online/offline and serial/parallel. In addition, we discuss other characteristics such as prediction and detection techniques, the characteristics used for prediction, interval type, etc. We also identify gaps in the state-of-the-art and discuss future research directions to enable and fully exploit the benefits of adaptable computing.

show abstract

Managing multi-configuration hardware via dynamic working set analysis

Abstract: Abstract

Cited by 96 publications

References 24 publications

Caracterização do perfil de consumo de recursos de programas binários utilizando a técnica DAMICORE

Caracterização do perfil de consumo de recursos de programas binários utilizando a técnica DAMICORE

Efficiency trends and limits from comprehensive microarchitectural adaptivity

A Survey of Phase Classification Techniques for Characterizing Variable Application Behavior

Contact Info

Product

Resources

About