Hessle-Free

Moazzemi, Kasra; Maity, Biswadip; Rahmani, Amir M.; Dutt, Nikil

doi:10.1145/3358203

Cited by 23 publications

(3 citation statements)

References 74 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Heterogeneous systems on chip (SoCs) combine the energy efficiency and performance of custom designs with the flexibility benefits of general-purpose cores. Domain-specific SoCs (DSSoCs), a subset of heterogeneous SoCs, are emerging examples that integrate general-purpose compute elements and hardware accelerators that target the commonly encountered tasks (i.e., computational kernels) in the target domain [1][2][3][4]. For example, a DSSoC for autonomous driving incorporates computer vision and deep learning accelerators, whereas a DSSoC for a 5G/6G communication domain accelerates signal processing operations such as a fast Fourier transform (FFT).…”

Section: Introductionmentioning

confidence: 99%

Theoretical Validation and Hardware Implementation of Dynamic Adaptive Scheduling for Heterogeneous Systems on Chip

Goksoy,

Hassan,

Krishnakumar

et al. 2023

JLPEA

View full text Add to dashboard Cite

Domain-specific systems on chip (DSSoCs) aim to narrow the gap between general-purpose processors and application-specific designs. CPU clusters enable programmability, whereas hardware accelerators tailored to the target domain minimize task execution times and power consumption. Traditional operating system (OS) schedulers can diminish the potential of DSSoCs, as their execution times can be orders of magnitude larger than the task execution time. To address this problem, we propose a dynamic adaptive scheduling (DAS) framework that combines the advantages of a fast, low-overhead scheduler and a sophisticated, high-performance scheduler with a larger overhead. We present a novel runtime classifier that chooses the better scheduler type as a function of the system workload, leading to improved system performance and energy-delay product (EDP). Experiments with five real-world streaming applications indicate that DAS consistently outperforms fast, low-overhead, and slow, sophisticated schedulers. DAS achieves a 1.29× speedup and a 45% lower EDP than the sophisticated scheduler under low data rates and a 1.28× speedup and a 37% lower EDP than the fast scheduler when the workload complexity increases. Furthermore, we demonstrate that the superior performance of the DAS framework also applies to hardware platforms, with up to a 48% and 52% reduction in the execution time and EDP, respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

Theoretical Validation and Hardware Implementation of Dynamic Adaptive Scheduling for Heterogeneous Systems on Chip

Goksoy,

Hassan,

Krishnakumar

et al. 2023

JLPEA

View full text Add to dashboard Cite

show abstract

“…Uma das técnicas de inteligência computacional utilizadas na tomada de decisão e seleção de conceitos variados é a Lógica Fuzzy (Arji et al, 2019;Dhunny et al, 2019;Droste et al, 2019), essa que é empregada em ferramentas como MatLab® para pesquisas cientificas, unida à Engenharia de Software surgem os frameworks que visam facilitar o uso de algoritmos como o citado para a geração de resultados em motores de processamento tal como é feito nos processos de fuzzificação e defuzzificação (Eslami Giski, Ebrahimzadeh, & Markechová, 2019;Magdalena, 2019;Ojha, Abraham, & Snášel, 2019). Development, v. 9, n. 9, e55995531, 2020 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i9.5531 Dessa forma o estudo dos métodos desenvolvidos em bibliotecas como Fuzzylite (Elleuch & Wali, 2019;Moazzemi et al, 2019;Sauerländer-Biebl et al, 2017) são de cunho investigativo visando abstrair a coerência dos métodos matemáticos empregados, esses que são de suma importância para validar contextos em pesquisas científicas como esta.…”

Section: Introductionunclassified

Análise da dispersão dos indicadores de desempenho para a tomada de decisão entre duas ferramentas de modelagem fuzzy

Silva¹,

Júnior²,

Parente³

et al. 2020

RSD

View full text Add to dashboard Cite

Analisar métodos de processamento de dados e garantir resultados coerentes é um trabalho árduo para os cientistas, o fato é que a otimização de um processo faz a diferença independente de qual área esteja sendo trabalhada. Existem ferramentas que consolidam estudos como esse, uma delas é o método comparativo utilizando análise de dispersão que foi utilizado na presente pesquisa ao trabalhar com o aplicativo Aventureiros Escolares que é o um dos objetos de estudo, com ele foi analisado o agente de inferências denominado Fuzzylite. O estudo de caso foi uma boa opção para a pesquisa ao analisar os métodos de inferência modelados no app e realizar uma comparação com o Toolbox Fuzzy do MatLab®, com isso o objetivo da presente pesquisa é analisar as duas ferramentas citadas e apontar qual delas teve o melhor desempenho utilizando a mesma modelagem Fuzzy. Por meio da análise de dispersão foi possível identificar um melhor resultado para o modelo empregado no Fuzzylite.

show abstract

“…As mentioned by Xun, Lei et al [16][17][18][19] current approaches for EML mainly focus on hardware or algorithm optimizations. Other research [20][21][22][23] has focused on optimization opportunities in either applications or hardware devices. However, it is important to explore other opportunities of the system as a whole and not just the application or devices based upon the diversity of computation on an embedded platform.…”

mentioning

confidence: 99%

Towards QoS-Based Embedded Machine Learning

et al. 2022

View full text Add to dashboard Cite

Due to various breakthroughs and advancements in machine learning and computer architectures, machine learning models are beginning to proliferate through embedded platforms. Some of these machine learning models cover a range of applications including computer vision, speech recognition, healthcare efficiency, industrial IoT, robotics and many more. However, there is a critical limitation in implementing ML algorithms efficiently on embedded platforms: the computational and memory expense of many machine learning models can make them unsuitable in resource-constrained environments. Therefore, to efficiently implement these memory-intensive and computationally expensive algorithms in an embedded computing environment, innovative resource management techniques are required at the hardware, software and system levels. To this end, we present a novel quality-of-service based resource allocation scheme that uses feedback control to adjust compute resources dynamically to cope with the varying and unpredictable workloads of ML applications while still maintaining an acceptable level of service to the user. To evaluate the feasibility of our approach we implemented a feedback control scheduling simulator that was used to analyze our framework under various simulated workloads. We also implemented our framework as a Linux kernel module running on a virtual machine as well as a Raspberry Pi 4 single board computer. Results illustrate that our approach was able to maintain a sufficient level of service without overloading the processor as well as providing an energy savings of almost 20% as compared to the native resource management in Linux.

show abstract

Hessle-Free

Cited by 23 publications

References 74 publications

Theoretical Validation and Hardware Implementation of Dynamic Adaptive Scheduling for Heterogeneous Systems on Chip

Theoretical Validation and Hardware Implementation of Dynamic Adaptive Scheduling for Heterogeneous Systems on Chip

Análise da dispersão dos indicadores de desempenho para a tomada de decisão entre duas ferramentas de modelagem fuzzy

Towards QoS-Based Embedded Machine Learning

Contact Info

Product

Resources

About