Pipelets: Self-Organizing Software Pipelines for Many-Core Architctures

Jahn, Janmartin; Henkel, Jörg

doi:10.7873/date.2013.308

Cited by 5 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…N apps is always equal or greater to 1 under the convention that there is an administrative application which is responsible for idle computational resources handling (see Section 3.1.3). If we define as R i the resources occupied by the i-th application then the following two equations are valid: (6) implying that that two applications cannot share the same computational resources (Eq. 5) and the total computational resources occupied by all applications are equal to the total resources of the system (Eq.…”

Section: Target Application Modelmentioning

confidence: 99%

“…Many-core architectures [2], [3], [4] are nowadays a reality, proposed as an effective computer organization to address the ever increasing user demands for higher performance, reliability and lower power consumption. The nature of applications' design is also evolving by incorporating dynamic characteristics such as high variability in workload, self-awareness and malleability, i.e seamless run-time adaptivity to available system resources [5], [6], [7].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Application-Arrival Rate Aware Distributed Run-Time Resource Management for Many-Core Computing Platforms

Tsoutsouras

Xydis

Soudris

2018

IEEE Trans. Multi-Scale Comp. Syst.

View full text Add to dashboard Cite

Modern many-core computing platforms execute a diverse set of dynamic workloads in the presence of varying application arrival rates. This inflicts strict requirements on run-time management to efficiently allocate system resources. On the way towards kilocore processor architectures, centralized resource management approaches will most probably form a severe performance bottleneck, thus focus has been turned to the study of Distributed Run-Time Resource Management (DRTRM) schemes. In this article we examine the behaviour of a DRTRM of dynamic applications with malleable characteristics against stressing incoming application interval rate scenarios, using Intel SCC as the target many-core system. We show that resource allocation is highly affected by application input rate and propose an application-arrival aware DRTRM framework implementing an effective admission control strategy by carefully utilizing voltage and frequency scaling on parts of its resource allocation infrastructure. Through extensive experimental evaluation, we quantitatively analyse the behavior of the introduced DRTRM scheme and show that it achieves up to 44% performance gains while consuming 31% less energy, in comparison to a state-of-art DRTRM solution. In comparison to a centralized RTRM, the respective metric values rise up to 62% and 45% performance and energy gains, respectively.

show abstract

Section: Target Application Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Application-Arrival Rate Aware Distributed Run-Time Resource Management for Many-Core Computing Platforms

Tsoutsouras

Xydis

Soudris

2018

IEEE Trans. Multi-Scale Comp. Syst.

View full text Add to dashboard Cite

show abstract

“…To adapt resource allocations autonomously at runtime without a controlling instances, we present Pipelets as selforganizing software pipelines [20]. A Pipelet is a stage of a software pipeline that perpetually monitors its resource requirements and starts to interact with other Pipelets to decide about adapting resource allocations by migrating to a different core.…”

Section: Self-organizing Resource Allocationmentioning

confidence: 99%

“…Compilers and frameworks must assist programmers and support resource allocation schemes so that parallel applications (e.g. software pipelines or Pipelets [20] can be created easily. Work on automated parallelization of programs concentrates on extracting parallelism from sequential applications.…”

Section: Research Challengesmentioning

confidence: 99%

See 1 more Smart Citation

Runtime resource allocation for software pipelines

Jahn

Kobbe

Pagani

et al. 2013

Proceedings of the 16th International Workshop on Software and Compilers for Embedded Systems

Self Cite

View full text Add to dashboard Cite

Systems continue to comprise a rapidly growing number of cores on a single chip to gain performance benefits from parallel processing. A key challenge is how their computational resources can be used efficiently, which depends to a large degree on how their resources are allocated to the applications. In this paper, we describe our current research for addressing this challenge and highlight current and upcoming hurdles that need to be addressed.

show abstract