Analyzing the Impact of CPU Pinning and Partial CPU Loads on Performance and Energy Efficiency

Podzimek, Andrej; Bulej, Lubomír; Chen, Lydia Y.; Binder, Walter; Tůma, Petr

doi:10.1109/ccgrid.2015.164

Cited by 40 publications

(33 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here we complement the results presented in our prior work [12] with a study of the overhead observed in the employed colocation solutions (KVM virtual machines and LXC containers) at partial loads.…”

Section: Introductionmentioning

confidence: 53%

“…The performance penalty is relatively strong for background loads around 0.5, yet almost negligible for both high and low background loads. 12 It appears to be weaker for LXC than for KVM, stronger for luindex and the Scala benchmarks and almost negligible for h2 and avrora. The latter correlates with the fact that ''per-thread'' is not always the foreground performance winner for avrora and h2, as shown in Fig.…”

Section: Background Load and Foreground Throughputmentioning

confidence: 95%

See 1 more Smart Citation

Robust partial-load experiments with Showstopper

Podzimek

Bulej

Chen

et al. 2016

Future Generation Computer Systems

Self Cite

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 53%

Section: Background Load and Foreground Throughputmentioning

confidence: 95%

Robust partial-load experiments with Showstopper

Podzimek

Bulej

Chen

et al. 2016

Future Generation Computer Systems

Self Cite

View full text Add to dashboard Cite

“…Prior work has demonstrated the importance of thread-to-core bindings in the overall performance of a parallelized application [8]. The task of discovering such optimal bindings is rather complex, given the structure of NUMA architectures [9].…”

Section: Related Work and Contributionsmentioning

confidence: 99%

“…Let also denote L * , the minimum makespan that can be achieved at the optimal assignments. Proposition 5.1 (Existence of Nash equilibria): Consider the load-balancing game characterized by the tuple I, A, {u i } i with a utility function defined by (8). Then, the set of pure Nash equilibria is non-empty, i.e., B NE = ∅.…”

Section: Convergence Analysismentioning

confidence: 99%

Learning-Based Dynamic Pinning of Parallelized Applications in Many-Core Systems

Chasparis

Rossbory

Janjić

et al. 2019

2019 27th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP)

View full text Add to dashboard Cite

Motivated by the need for adaptive, secure and responsive scheduling in a great range of computing applications, including human-centered and time-critical applications, this paper proposes a scheduling framework that seamlessly adds resource-awareness to any parallel application. In particular, we introduce a learning-based framework for dynamic placement of parallel threads to Non-Uniform Memory Access (NUMA) architectures. Decisions are taken independently by each thread in a decentralized fashion that significantly reduces computational complexity. The advantage of the proposed learning scheme is the ability to easily incorporate any multi-objective criterion and easily adapt to performance variations during runtime. Under the multi-objective criterion of maximizing total completed instructions per second (i.e., both computational and memory-access instructions), we provide analytical guarantees with respect to the expected performance of the parallel application. We also compare the performance of the proposed scheme with the Linux operating system scheduler in an extensive set of applications, including both computationally and memory intensive ones. We have observed that performance improvement could be significant especially under limited availability of resources and under irregular memory-access patterns.

show abstract

“…We use a linear performance function: on each machine, the influence a type t has on type t performance is a product of the load of type t and a coefficient α t ,t . The coefficient α t ,t describes how compatible t load is with t performance (the coefficient is similar to interference/affinity metrics proposed in [13,20]). Low values (0 ≤ α t ,t < 1) correspond with compatible types (e.g.…”

Section: Introductionmentioning

confidence: 99%

Optimizing Egalitarian Performance in the Side-Effects Model of Colocation for Data Center Resource Management

Pascual

Rządca

2017

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In data centers, up to dozens of tasks are colocated on a single physical machine. Machines are used more efficiently, but tasks' performance deteriorates, as colocated tasks compete for shared resources. As tasks are heterogeneous, the resulting performance dependencies are complex. In our previous work [18] we proposed a new combinatorial optimization model that uses two parameters of a task -its size and its type -to characterize how a task influences the performance of other tasks allocated to the same machine. In this paper, we study the egalitarian optimization goal: maximizing the worstoff performance. This problem generalizes the classic makespan minimization on multiple processors (P||C max ). We prove that polynomially-solvable variants of P||C max are NP-hard and hard to approximate when the number of types is not constant. For a constant number of types, we propose a PTAS, a fast approximation algorithm, and a series of heuristics. We simulate the algorithms on instances derived from a trace of one of Google clusters. Algorithms aware of jobs' types lead to better performance compared to algorithms solving P||C max . The notion of type enables us to model degeneration of performance caused by colocation using standard combinatorial optimization methods. Types add a layer of additional complexity. However, our results -approximation algorithms and good average-case performance -show that types can be handled efficiently.

show abstract

Analyzing the Impact of CPU Pinning and Partial CPU Loads on Performance and Energy Efficiency

Cited by 40 publications

References 39 publications

Robust partial-load experiments with Showstopper

Robust partial-load experiments with Showstopper

Learning-Based Dynamic Pinning of Parallelized Applications in Many-Core Systems

Optimizing Egalitarian Performance in the Side-Effects Model of Colocation for Data Center Resource Management

Contact Info

Product

Resources

About