LAANT: A library to automatically optimize EDP for OpenMP applications

Lorenzon, Arthur F.; Souza, Jeckson Dellagostin; Beck, Antonio Carlos Schneider

doi:10.23919/date.2017.7927176

Cited by 24 publications

(7 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Energy Efficiency: Improving energy efficiency through resource management policies has been investigated as well, in studies such as OpenMPE [1], Benedict et al [3], and LAANT [14]. In the former, an OpenMP extension designed to improve energy management is proposed.…”

Section: Related Workmentioning

confidence: 99%

“…In [3], the authors propose an energy prediction mechanism for OpenMP applications using a Random Forest Modeling approach. LAANT [14] is a library that aims at optimizing the EDP metric. The study conducted in Porterfield et al [18] similarly proposes a system to automatically adjust the number of threads based on on-line measurements of system resource usage.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Enhancing Resource Management through Prediction-based Policies

Navarro,

Lorenzon,

Ayguadé

et al. 2020

Preprint

View full text Add to dashboard Cite

Task-based programming models are emerging as a promising alternative to make the most of multi-/many-core systems. These programming models rely on runtime systems, and their goal is to improve application performance by properly scheduling application tasks to cores. Additionally, these runtime systems offer policies to cope with application phases that lack in parallelism to fill all cores. However, these policies are usually static and favor either performance or energy efficiency. In this paper, we have extended a task-based runtime system with a lightweight monitoring and prediction infrastructure that dynamically predicts the optimal number of cores required for each application phase, thus improving both performance and energy efficiency. Through the execution of several benchmarks in multi-/many-core systems, we show that our prediction-based policies have competitive performance while improving energy efficiency when compared to state of the art policies.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Enhancing Resource Management through Prediction-based Policies

Navarro,

Lorenzon,

Ayguadé

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Por outro lado, os processadores multicore com vários núcleos do tipo superescalar exploram o paralelismo no nível de instruc ¸ões, threads e dados (nestes, porém, em uma escala bem menor que as GPUs, através de instruc ¸ões SIMD -Single Instruction Multiple Data). Comparado com as GPUs, as CPUs -que possuem unidades de controle complexas e grandes caches [Khorasani et al 2014] -geralmente são melhores em realizar tarefas que demandam baixa latência, e que não possuem níveis de paralelismo de dados amplamente disponíveis para serem explorados [Lorenzon et al 2017, Lorenzon andBeck Filho 2019].…”

Section: Introduc ¸ãOunclassified

Decidindo entre GPU e CPU para Processar Grafos com Base em Métricas de Alto Nível

Moori

Rocha²,

Schwarzrock³

et al. 2022

Anais Do XXIII Simpósio Em Sistemas Computacionais De Alto Desempenho (SSCAD 2022)

View full text Add to dashboard Cite

Apesar dos avanços nas GPUs modernas ter acelerado a execução de aplicações que processam grandes quantidades de dados, acelerar o processamento de grafos nesses sistemas não é uma tarefa trivial: aplicações de grafos são caracterizadas por alto volume de acesso irregular à memória e que varia com a estrutura dos grafos, fazendo com que muitas vezes elas não alcancem seus picos de performance quando executadas em GPUs. Nesses casos, a execução em CPU é mais adequada. Felizmente, as estruturas dos grafos podem ser identificadas por meio de métricas de alto nível (e.g., diâmetro e coeficiente médio de clusterização), e elas podem auxiliar o projetista na tomada de decisão de onde executar a aplicação, se em GPU ou em CPU. Neste trabalho, nós propomos uma metodologia que usa essas características de alto nível em uma Regressão Linear para auxiliar na tomada de decisão de onde processar os grafos, em GPU ou CPU. Assim, sempre que um novo grafo precisar ser processado, a decisão de onde processá-lo será tomada com base nessas métricas, evitando qualquer execução adicional dos algoritmos de grafos. Nosso resultados experimentais, considerando 1 GPU e 2 CPUs, mostram que as métricas mais relevantes dos grafos variam conforme os algoritmo e as máquinas onde esses grafos serão executados. Nossa proposta apresentou uma acurácia média de 85% quando aplicado uma Regressão Linear com as características mais relevantes.

show abstract

“…However, graph applications, because of their irregular structure and larger dimension, are intrinsically communication/memory‐bound, which means that they tend to use the CPU less when compared to the average parallel application 1,10 . Given that, graph applications executions are even more affected by issues that tend to limit the scalability of parallel applications related to both hardware (e.g., saturation of execution units and communication bus) and software (e.g., data synchronization and concurrent accesses to shared memory) 11‐13 . This added to the fact that HPC multiprocessors have their memory hierarchy getting more complex, may indicate that executing all graph applications with the maximum number of threads, which is the default behavior, may not be efficient 14,15 .…”

Section: Introductionmentioning

confidence: 99%

Improving the efficiency of graph algorithm executions on high‐performance computing

Moori

Rocha

Schwarzrock

et al. 2022

Concurrency and Computation

View full text Add to dashboard Cite

The growing need for extracting information from large graphs has been pushing the development of parallel graph algorithms. However, the highly irregular structure of the real-world graphs limits the performance and energy improvements of graph applications. In this paper, we show that, in most cases, using all the available cores of the multiprocessor is not the best option in terms of the aforementioned non-functional requirements. Based on that, we propose GraphKat, a framework that enables the simultaneous processing of several algorithms/graphs instead of executing them serially (i.e., one after another), increasing efficiency in terms of performance and energy.GraphKat works in two steps: (i) it characterizes the graph applications with a specific number of threads based on their efficiency levels; and (ii) it defines the execution order of all graph applications in the target system. Experimental results on three multicore processors (Intel and AMD) show that GraphKat improves the overall system's efficiency related to performance (up to 434.26×) and energy-saving (up to 245.21×), and reduces the graph applications' execution time (up to 17.70×) and energy consumption (up to 6.64×) compared to the default execution of parallel applications on HPC systems.

show abstract

LAANT: A library to automatically optimize EDP for OpenMP applications

Cited by 24 publications

References 10 publications

Enhancing Resource Management through Prediction-based Policies

Enhancing Resource Management through Prediction-based Policies

Decidindo entre GPU e CPU para Processar Grafos com Base em Métricas de Alto Nível

Improving the efficiency of graph algorithm executions on high‐performance computing

Contact Info

Product

Resources

About