Person: TOPCUOĞLU, HALUK RAHMİ
Loading...
Email Address
Birth Date
Research Projects
Organizational Units
Job Title
Last Name
TOPCUOĞLU
First Name
HALUK RAHMİ
Name
31 results
Search Results
Now showing 1 - 10 of 31
Publication Metadata only A task scheduling algorithm for arbitrarily-connected processors with awareness of link contention(SPRINGER, 2006) ALKAYA, ALİ FUAT; Alkaya, Ali Fuat; Topcuoglu, Haluk RahmiIn this paper, we present a new task scheduling algorithm, called Contention-Aware Scheduling (CAS) algorithm, with the objective of delivering good quality of schedules in low running-time by considering contention on links of arbitrarily-connected, heterogeneous processors. The CAS algorithm schedules tasks on processors and messages on links by considering the earliest finish time attribute with the virtual cut-through (VCT) or the store- and-forward (SAF) switching. There are three types of CAS algorithm presented in this paper, which differ in ordering the messages from immediate predecessor tasks. As part of the experimental study, the performance of the CAS algorithm is compared with two well-known APN (arbitrary processor network) scheduling algorithms. Experiments on the results of the synthetic benchmarks and the task graphs of the well-known problems clearly show that our CAS algorithm outperforms the related work with respect to performance (given in normalized schedule length) and cost (given in running time) to generate output schedules.Publication Open Access Quantifying the impact of data replication on error propagation(2022-09-01) ÖZTÜRK, ZUHAL; TOPCUOĞLU, HALUK RAHMİ; ÖZTÜRK Z., TOPCUOĞLU H. R. , Kandemir M. T.Various technological developments in the microprocessor world make modern computing systems more vulnerable to soft errors than in the past, and consequently fault tolerance techniques are becoming increasingly important in various application domains. While in general fault tolerance methods are known to achieve high levels of reliability, they can also introduce significant performance, energy, and memory overheads, which can be reduced by employing such techniques selectively, as opposed to indiscriminately. Data Replication is used to prevent error propagation across hardware components and application program data structures by replicating application program\"s data. When using data replication, many factors need to be taken into account, including which data structures/elements to replicate, how many times to replicate a given data element, and which threads to protect (in a multithreaded application). These and similar factors define what can be termed as \"replication space\". This study defines a replication space, and systematically explores protection techniques of various strengths/degrees, quantifying their impacts on memory consumption, performance, and error propagation. Our experimental analysis reveals that different degrees of protection levels bring different outcomes based on the application specifics. In particular, while error propagation is limited, to a certain extent, when employing data replication in multithreaded applications where the thread do not communicate/share data much, the speed of error propagation across threads can be quite fast in applications where threads are more tightly coupled. Additionally, our results indicate that in certain cases where error propagation is low, the effect of data replication on error propagation can be negligible.Publication Metadata only Thread vulnerability in parallel applications(ACADEMIC PRESS INC ELSEVIER SCIENCE, 2012) TOPCUOĞLU, HALUK RAHMİ; Oz, Isil; Topcuoglu, Haluk Rahmi; Kandemir, Mahmut; Tosun, OguzContinuously reducing transistor sizes and aggressive low power operating modes employed by modern architectures tend to increase transient error rates. Concurrently, multicore machines are dominating the architectural spectrum today in various application domains. These two trends require a fresh look at resiliency of multithreaded applications against transient errors from a software perspective. In this paper, we propose and evaluate a new metric called the Thread Vulnerability Factor (TVF). A distinguishing characteristic of TVF is that its calculation for a given thread (which is typically one of the threads of a multithreaded application) does not depend on its code alone, but also on the codes of the threads that share resources and data with that thread. As a result, we decompose TVF of a thread into two complementary parts: local and remote. While the former captures the TVF induced by the code of the target thread, the latter represents the vulnerability impact of the threads that interact with the target thread. We quantify the local and remote TVF values for three architectural components (register file, ALUs, and caches) using a set of ten multithreaded applications from the Parsec and Splash-2 benchmark suites. Our experimental evaluation shows that TVF values tend to increase as the number of cores increases, which means the system becomes more vulnerable as the core count rises. We further discuss how TVF metric can be employed to explore performance-reliability tradeoffs in multicores. Reliability-based analysis of compiler optimizations and redundancy-based fault tolerance are also mentioned as potential usages of our TVF metric. (C) 2012 Elsevier Inc. All rights reserved.Publication Metadata only A Hybrid Evolutionary Algorithm for Solving the Register Allocation Problem(Springer Verlag, 2004) BOZ, BETÜL; Demiroz B., Topcuoglu H., Kandemir M.One of the strong impacts on runtime performance of a given code is the register allocation phase of compilation. It is crucial to provide aggressive and sophisticated register allocators for the embedded devices, where the excessive compilation time is tolerated because of its off-line nature. In this paper, we present a hybrid evolutionary algorithm for register allocation problem that combines genetic algorithms with a local search technique. Our algorithm exploits a novel, highly-specialized crossover operator that considers domain-specific information. Computational experiments performed on various synthetic benchmarks prove that our method significantly outperform the state-of-the-art algorithms with respect to all given comparison metrics on solution quality. © Springer-Verlag 2004.Publication Metadata only Neural network based multi-objective evolutionary algorithm for dynamic workflow scheduling in cloud computing(ELSEVIER, 2020) TOPCUOĞLU, HALUK RAHMİ; Ismayilov, Goshgar; Topcuoglu, Haluk RahmiWorkflow scheduling is a largely studied research topic in cloud computing, which targets to utilize cloud resources for workflow tasks by considering the objectives specified in QoS. In this paper, we model dynamic workflow scheduling problem as a dynamic multi-objective optimization problem (DMOP) where the source of dynamism is based on both resource failures and the number of objectives which may change over time. Software faults and/or hardware faults may cause the first type of dynamism. On the other hand, confronting real-life scenarios in cloud computing may change number of objectives at runtime during the execution of a workflow. In this study, we propose a prediction based dynamic multi-objective evolutionary algorithm, called NN-DNSGA-II algorithm, by incorporating artificial neural network with the NSGA-II algorithm. Additionally, five leading non-prediction based dynamic algorithms from the literature are adapted for the dynamic workflow scheduling problem. Scheduling solutions are found by the consideration of six objectives: minimization of makespan, cost, energy and degree of imbalance; and maximization of reliability and utilization. The empirical study based on real-world applications from Pegasus workflow management system reveals that our NN-DNSGA-II algorithm significantly outperforms the other alternatives in most cases with respect to metrics used for DMOPs with unknown true Pareto-optimal front, including the number of non-dominated solutions, Schott's spacing and Hypervolume indicator. (C) 2019 Elsevier B.V. All rights reserved.Publication Open Access Network Congestion Aware Multiobjective Task Scheduling in Heterogeneous Fog Environments(2023-01-01) TOPCUOĞLU, HALUK RAHMİ; Altin L., TOPCUOĞLU H. R., Gurgen F. S.Task scheduling on fog environments surges new challenges compared to scheduling on conventional cloud computing. Various levels of heterogeneity and dynamism cause task scheduling problem is more challenging for fog computing. In this study, we present a multiobjective task scheduling model with a total of five objectives and propose a multiobjective multirank (MOMRank) scheduling algorithm for fog computing. The performance of the proposed strategy is assessed with well-known multiobjective metaheuristics [the nondominated sorting genetic algorithm II (NSGA-II) and the Strength Pareto Evolutionary Algorithm 2 (SPEA2)] and a widely used algorithm from the literature, the multiobjective heterogeneous earliest finish time (MOHEFT) algorithm using three common multiobjective metrics. Additionally, we incorporate two task clustering mechanisms to the algorithms in order to improve data transmissions on interconnection networks. Results of empirical evaluations given in performance profiles over all problem instances validate significance of both our algorithm and the integrated extensions for diminishing data transfer costs.Publication Open Access Performance-effective and low-complexity task scheduling for heterogeneous computing(IEEE COMPUTER SOC, 2002-03) TOPCUOĞLU, HALUK RAHMİ; Topcuoglu, H; Hariri, S; Wu, MYEfficient application scheduling is critical for achieving high performance in heterogeneous computing environments. The application scheduling problem has been shown to be NP-complete in general cases as well as in several restricted cases. Because of its key importance, this problem has been extensively studied and various algorithms have been proposed in the literature which are mainly for systems with homogeneous processors. Although there are a few algorithms in the literature for heterogeneous processors, they usually require significantly high scheduling costs and they may not deliver good quality schedules with lower costs. In this paper, we present two novel scheduling algorithms for a bounded number of heterogeneous processors with an objective to simultaneously meet high performance and fast scheduling time, which are called the Heterogeneous Earliest-Finish-Time (HEFT) algorithm and the Critical-Path-on-a-Processor (CPOP) algorithm. The HEFT algorithm selects the task with the highest upward rank value at each step and assigns the selected task to the processor, which minimizes its earliest finish time with an insertion-based approach. On the other hand, the CPOP algorithm uses the summation of upward and downward rank values for prioritizing tasks. Another difference is in the processor selection phase, which schedules the critical tasks onto the processor that minimizes the total execution time of the critical tasks. In order to provide a robust and unbiased comparison with the related work, a parametric graph generator was designed to generate weighted directed acyclic graphs with various characteristics. The comparison study, based on both randomly generated graphs and the graphs of some real applications, shows that our scheduling algorithms significantly surpass previous approaches in terms of both quality and cost of schedules, which are mainly presented with schedule length ratio, speedup, frequency of best results, and average scheduling time metrics.Publication Metadata only Scheduling opportunities for asymmetrically reliable caches(ACADEMIC PRESS INC ELSEVIER SCIENCE, 2019) TOPCUOĞLU, HALUK RAHMİ; Arslan, Sanem; Topcuoglu, Haluk Rahmi; Kandemir, Mahmut Taylan; Tosun, OguzModern systems become more vulnerable to soft errors with technology scaling. Providing fault tolerance strategies on all structures in a system may lead to high energy consumption. Our framework with asymmetrically reliable caches with at least one protected core and several unprotected cores dynamically assigns the software threads executing critical code fragments to the protected core(s) with the FCFS-based algorithm. The framework can provide good reliability, performance, and power consumption trade-offs compared with the fully protected and unprotected systems. However, FCFS-based scheduling algorithm may degrade the system performance and unfairly slow down applications for some workloads. In this paper, a set of scheduling algorithms is proposed to improve both the system performance and fairness perspectives. Various static priority techniques that require preliminary information about the applications (such as their execution order, cache usage, number of requests sent to the protected core(s), and total burst time spent on the protected core(s)) are implemented and evaluated. On the other hand, dynamic priority techniques that target to equalize the total time spent of applications on the protected core(s) or the progress of the applications' requests are presented. Extensive evaluations using multi application workloads validate significant improvements of our static and dynamic priority scheduling techniques on system performance and fairness over the FCFS algorithm. (C) 2019 Elsevier Inc. All rights reserved.Publication Metadata only Impact of sensor-based change detection schemes on the performance of evolutionary dynamic optimization techniques(SPRINGER, 2018) TOPCUOĞLU, HALUK RAHMİ; Altin, Lokman; Topcuoglu, Haluk RahmiEvolutionary algorithms are among the most common techniques developed to address dynamic optimization problems. They either assume that changes in the environment are known a priori, especially for some benchmark problems, or detect these changes. On the other hand, detecting the points in time where a change occurs in the landscape is a critical issue. In this paper, we investigate the performance evaluation of various sensor-based detection schemes on the moving peaks benchmark and the dynamic knapsack problem. Our empirical study validates the performance of the sensor-based detection schemes considered, by using the average rate of correctly identified changes and number of sensors invoked to detect a change. We also propose a new mechanism to evaluate the capability of the detection schemes for determining severity of changes. Additionally, a novel hybrid approach is proposed by integrating the change detection schemes with evolutionary dynamic optimization algorithms in order to set algorithm-specific parameters dynamically. The experimental evaluation validates that our extensions outperform the reference algorithms for various characteristics of dynamism.Publication Metadata only Performance evaluation of evolutionary heuristics in dynamic environments(SPRINGER, 2012) TOPCUOĞLU, HALUK RAHMİ; Ayvaz, Demet; Topcuoglu, Haluk Rahmi; Gurgen, FikretIn recent years, there has been a growing interest in applying genetic algorithms to dynamic optimization problems. In this study, we present an extensive performance evaluation and comparison of 13 leading evolutionary algorithms with different characteristics on a common platform by using the moving peaks benchmark and by varying a set of problem parameters including shift length, change frequency, correlation value and number of peaks in the landscape. In order to compare solution quality or the efficiency of algorithms, the results are reported in terms of both offline error metric and dissimilarity factor, our novel comparison metric presented in this paper, which is based on signal similarity. Computational effort of each algorithm is reported in terms of average number of fitness evaluations and the average execution time. Our experimental evaluation indicates that the hybrid methods outperform the related work with respect to quality of solutions for various parameters of the given benchmark problem. Specifically, hybrid methods provide up to 24% improvement with respect to offline error and up to 30% improvement with respect to dissimilarity factor by requiring more computational effort than other methods.