Runtime Estimation Research Articles

Local Source Based CFD Modeling of Effusion Cooling Holes: Validation and Application to an Actual Combustor Test Case

State-of-the-art liner cooling technology for modern combustion chambers is represented by effusion cooling (or full-coverage film cooling). Effusion is a very efficient cooling strategy typically based on the use of several inclined small diameter cylindrical holes, where liner temperature is controlled by the combined protective effect of coolant film and heat removal through forced convection inside each hole. A CFD-based thermal analysis of such components implies a significant computational cost if the cooling holes are included in the simulations; therefore many efforts have been made to develop lower order approaches aiming at reducing the number of mesh elements. The simplest approach models the set of holes as a uniform coolant injection, but it does not allow an accurate assessment of the interaction between hot gas and coolant. Therefore higher order models have been developed, such as those based on localized mass sources in the region of hole discharge. The model presented in this paper replaces the effusion hole with a mass sink on the cold side of the plate, a mass source on the hot side, whereas convective cooling within the perforation is accounted for with a heat sink. The innovative aspect of the work is represented by the automatic calculation of the mass flow through each hole, obtained by a run time estimation of isentropic mass flow with probe points, while the discharge coefficients are calculated at run time through an in-house developed correlation. In the same manner, the heat sink is calculated from a Nusselt number correlation available in literature for short length holes. The methodology has been applied to experimental test cases of effusion cooling plates and compared to numerical results obtained through a CFD analysis including the cooling holes, showing a good agreement. A comparison between numerical results and experimental data was performed on an actual combustor as well, in order to prove the feasibility of the procedure.

Job scheduling with adjusted runtime estimates on production supercomputers

The estimate of a parallel job’s running time (walltime) is an important attribute used by resource managers and job schedulers in various scenarios, such as backfilling and short-job-first scheduling. This value is provided by the user, however, and has been repeatedly shown to be inaccurate. We studied the workload characteristic based on a large amount of historical data (over 275,000 jobs in two and a half years) from a production leadership-class computer. Based on that study, we proposed a set of walltime adjustment schemes producing more accurate estimates. To ensure the utility of these schemes on production systems, we analyzed their potential impact in scheduling and evaluated the schemes with an event-driven simulator. Our experimental results show that our method can achieve not only better overall estimation accuracy but also improved overall system performance. Specifically, the average estimation accuracy of the tested workload can be improved by up to 35%, and the system performance in terms of average waiting time and weighted average waiting time can be improved by up to 22% and 28%, respectively.

Online cost-efficient scheduling of deadline-constrained workloads on hybrid clouds

Cloud computing has found broad acceptance in both industry and research, with public cloud offerings now often used in conjunction with privately owned infrastructure. Technical aspects such as the impact of network latency, bandwidth constraints, data confidentiality and security, as well as economic aspects such as sunk costs and price uncertainty are key drivers towards the adoption of such a hybrid cloud model. The use of hybrid clouds introduces the need to determine which workloads are to be outsourced, and to what cloud provider. These decisions should minimize the cost of running a partition of the total workload on one or multiple public cloud providers while taking into account the application requirements such as deadline constraints and data requirements. The variety of cost factors, pricing models and cloud provider offerings to consider, further calls for an automated scheduling approach in hybrid clouds. In this work, we tackle this problem by proposing a set of algorithms to cost-efficiently schedule the deadline-constrained bag-of-tasks applications on both public cloud providers and private infrastructure. Our algorithms take into account both computational and data transfer costs as well as network bandwidth constraints. We evaluate their performance in a realistic setting with respect to cost savings, deadlines met and computational efficiency, and investigate the impact of errors in runtime estimates on these performance metrics.

Multiple Workflow Scheduling Strategies with User Run Time Estimates on a Grid

In this paper, we present an experimental study of deterministic non-preemptive multiple workflow scheduling strategies on a Grid. We distinguish twenty five strategies depending on the type and amount of information they require. We analyze scheduling strategies that consist of two and four stages: labeling, adaptive allocation, prioritization, and parallel machine scheduling. We apply these strategies in the context of executing the Cybershake, Epigenomics, Genome, Inspiral, LIGO, Montage, and SIPHT workflows applications. In order to provide performance comparison, we performed a joint analysis considering three metrics. A case study is given and corresponding results indicate that well known DAG scheduling algorithms designed for single DAG and single machine settings are not well suited for Grid scheduling scenarios, where user run time estimates are available. We show that the proposed new strategies outperform other strategies in terms of approximation factor, mean critical path waiting time, and critical path slowdown. The robustness of these strategies is also discussed.

Exploiting the Burstiness of Intermediate-Quality Wireless Links

We address the challenge of link estimation and routing over highly dynamic links, thats is, bursty links that rapidly shift between reliable and unreliable periods of transmissions. Based on significant empirical evidence of over 100,000 transmissions over each link in 802.15.4 and 802.11 testbeds, we propose two metrics, expected future transmissions (EFT) and MAC3, for runtime estimation of bursty wireless links. We introduce a bursty link estimator (BLE) that based on these two metrics, accurately estimates bursty links in the network rendering them available for data transmissions. Finally, we present bursty routing extensions (BRE): an adaptive routing strategy that uses BLE for forwarding packets over bursty links if they offer better routing progress than long-term stable links. Our evaluation, comprising experimental data from widely used IEEE 802.15.4-based testbeds, reveals an average of 19% and a maximum of 42% reduction in the number of transmissions when routing over long-range bursty links typically ignored by routing protocols. Additionally, we show that both BLE and BRE are not tied to any specific routing protocol and integrate seamlessly with existing routing protocols and link estimators.

Thermalization in Nature and on a Quantum Computer

In this work, we show how Gibbs or thermal states appear dynamically in closed quantum many-body systems, building on the program of dynamical typicality. We introduce a novel perturbation theorem for physically relevant weak system-bath couplings that is applicable even in the thermodynamic limit. We identify conditions under which thermalization happens and discuss the underlying physics. Based on these results, we also present a fully general quantum algorithm for preparing Gibbs states on a quantum computer with a certified runtime and error bound. This complements quantum Metropolis algorithms, which are expected to be efficient but have no known runtime estimates and only work for local Hamiltonians.

Approximating the set of local minima in partial RNA folding landscapes

We study a stochastic method for approximating the set of local minima in partial RNA folding landscapes associated with a bounded-distance neighbourhood of folding conformations. The conformations are limited to RNA secondary structures without pseudoknots. The method aims at exploring partial energy landscapes pL induced by folding simulations and their underlying neighbourhood relations. It combines an approximation of the number of local optima devised by Garnier and Kallel (2002) with a run-time estimation for identifying sets of local optima established by Reeves and Eremeev (2004). The method is tested on nine sequences of length between 50 nt and 400 nt, which allows us to compare the results with data generated by RNAsubopt and subsequent barrier tree calculations. On the nine sequences, the method captures on average 92% of local minima with settings designed for a target of 95%. The run-time of the heuristic can be estimated by O(n(2)Dνlnν), where n is the sequence length, ν is the number of local minima in the partial landscape pL under consideration and D is the maximum number of steepest descent steps in attraction basins associated with pL.

Coordinated rescheduling of Bag‐of‐Tasks for executions on multiple resource providers

SUMMARYMetaschedulers can distribute parts of a Bag‐of‐Tasks (BoT) application among various resource providers in order to speed up its execution. The expected completion time of the user application is then calculated based on the run‐time estimates of all applications running and waiting for resources. However, because of inaccurate run time estimates, initial schedules are not those that provide users with the earliest completion time. These estimates increase the time distance between the first and last tasks of a BoT application, which increases average user response time, especially in multi‐provider environments. This paper proposes a coordinated rescheduling algorithm to handle inaccurate run‐time estimates when executing BoT applications in multi‐provider environments. The coordinated rescheduling defines which tasks can have start time updated based on the expected completion time of the entire BoT application. We have also evaluated the impact of system‐generated run‐time estimates to schedule BoT applications on multiple providers. We performed experiments using simulations and a real distributed platform, Grid'5000. From our experiments, we obtained reductions of up to 5 and 10% for response time and slowdown metrics, respectively, by using coordinated rescheduling over a traditional rescheduling solution. Moreover, coordinated rescheduling requires little modification of existing scheduling systems. System‐generated predictions, on the other hand, are more complex to be deployed and may not reduce response times as much as coordinated rescheduling. Copyright © 2011 John Wiley & Sons, Ltd.

Using accurate AIC-based performance models to improve the scheduling of parallel applications

Predictions based on analytical performance models can be used on efficient scheduling policies in order to select adequate resources for an optimal execution in terms of throughput and response time. However, developing accurate analytical models of parallel applications is a hard issue. The TIA (Tools for Instrumenting and Analysis) modeling framework provides an easy to use modeling method for obtaining analytical models of MPI applications. This method is based on modeling selection techniques and, in particular, on Akaike’s information criterion (AIC). In this paper, first the AIC-based performance model of the HPL benchmark is obtained using the TIA modeling framework. Then the use of this model for assessing the runtime estimation on different backfilling policies is analyzed in the GridSim simulator. The behavior of these simulations is compared with the equivalent simulations based on the theoretical model of the HPL provided by its developers.

Job Allocation Strategies with User Run Time Estimates for Online Scheduling in Hierarchical Grids

We address non-preemptive non-clairvoyant online scheduling of parallel jobs on a Grid. We consider a Grid scheduling model with two stages. At the first stage, jobs are allocated to a suitable Grid site, while at the second stage, local scheduling is independently applied to each site. We analyze allocation strategies depending on the type and amount of information they require. We conduct a comprehensive performance evaluation study using simulation and demonstrate that our strategies perform well with respect to several metrics that reflect both user- and system-centric goals. Unfortunately, user run time estimates and information on local schedules does not help to significantly improve the outcome of the allocation strategies. When examining the overall Grid performance based on real data, we determined that an appropriate distribution of job processor requirements over the Grid has a higher performance than an allocation of jobs based on user run time estimates and information on local schedules. In general, our experiments showed that rather simple schedulers with minimal information requirements can provide a good performance.

Wi-Fi 기반의 차량과 기지국간 통신에서 효과적인 AP 선택에 관한 연구

Wi-Fi based vehicle-to-infrastructure (V2I) communication is an emerging solution to improve the safety, traffic efficiency, and comfort of passengers. However, due to the high mobility of vehicles and the limited coverage of Wi-Fi APs, the V2I system may suffer from frequent handoffs although roadside APs can support cost effective Internet connectivity. Such problem of V2I systems can be overcome with Mobile AP (MAP) platform. The MAPs yield longer service duration by moving along with vehicles, yet they provide a lower link capacities than the roadside APs. In this paper, we propose a new association control mechanism that effectively determines whether the vehicle will select a fixed roadside-AP or a nearby MAP in mobile vehicular network environments. We consider both the achievable link bandwidth and available connection duration as a selection criterion and provide their run-time estimation method. Extensive simulation using real traces show significant performance improvements.

Enhanced sampling and applications in protein folding in explicit solvent.

We report a single-copy tempering method for simulating large complex systems. In a generalized ensemble, the method uses runtime estimate of the thermal average energy computed from a novel integral identity to guide a continuous temperature-space random walk. We first validated the method in a two-dimensional Ising model and a Lennard-Jones liquid system. It was then applied to folding of three small proteins, trpzip2, trp-cage, and villin headpiece in explicit solvent. Within 0.5-1 microsecond, all three systems were reversibly folded into atomic accuracy: the alpha carbon root mean square deviations of the best folded conformations from the native states were 0.2, 0.4, and 0.4 A, for trpzip2, trp-cage, and villin headpiece, respectively.

An Improved Smoothed $\ell^0$ Approximation Algorithm for Sparse Representation

l <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> norm based algorithms have numerous potential applications where a sparse signal is recovered from a small number of measurements. The direct l <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> norm optimization problem is NP-hard. In this paper we work with the the smoothed l <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> (SL0) approximation algorithm for sparse representation. We give an upper bound on the run-time estimation error. This upper bound is tighter than the previously known bound. Subsequently, we develop a reliable stopping criterion. This criterion is helpful in avoiding the problems due to the underlying discontinuities of the l <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sup> cost function. Furthermore, we propose an alternative optimization strategy, which results in a Newton like algorithm.

Model-driven autotuning of sparse matrix-vector multiply on GPUs

We present a performance model-driven framework for automated performance tuning (autotuning) of sparse matrix-vector multiply (SpMV) on systems accelerated by graphics processing units (GPU). Our study consists of two parts. First, we describe several carefully hand-tuned SpMV implementations for GPUs, identifying key GPU-specific performance limitations, enhancements, and tuning opportunities. These implementations, which include variants on classical blocked compressed sparse row (BCSR) and blocked ELLPACK (BELLPACK) storage formats, match or exceed state-of-the-art implementations. For instance, our best BELLPACK implementation achieves up to 29.0 Gflop/s in single-precision and 15.7 Gflop/s in double-precision on the NVIDIA T10P multiprocessor (C1060), enhancing prior state-of-the-art unblocked implementations (Bell and Garland, 2009) by up to 1.8× and 1.5× for single-and double-precision respectively. However, achieving this level of performance requires input matrix-dependent parameter tuning. Thus, in the second part of this study, we develop a performance model that can guide tuning. Like prior autotuning models for CPUs (e.g., Im, Yelick, and Vuduc, 2004), this model requires offline measurements and run-time estimation, but more directly models the structure of multithreaded vector processors like GPUs. We show that our model can identify the implementations that achieve within 15% of those found through exhaustive search.

An Integrated Approach to Locality-Conscious Processor Allocation and Scheduling of Mixed-Parallel Applications

Complex parallel applications can often be modeled as directed acyclic graphs of coarse-grained application tasks with dependences. These applications exhibit both task and data parallelism, and combining these two (also called mixed parallelism) has been shown to be an effective model for their execution. In this paper, we present an algorithm to compute the appropriate mix of task and data parallelism required to minimize the parallel completion time (makespan) of these applications. In other words, our algorithm determines the set of tasks that should be run concurrently and the number of processors to be allocated to each task. The processor allocation and scheduling decisions are made in an integrated manner and are based on several factors such as the structure of the task graph, the runtime estimates and scalability characteristics of the tasks, and the intertask data communication volumes. A locality-conscious scheduling strategy is used to improve intertask data reuse. Evaluation through simulations and actual executions of task graphs derived from real applications and synthetic graphs shows that our algorithm consistently generates schedules with a lower makespan as compared to Critical Path Reduction (CPR) and Critical Path and Allocation (CPA), two previously proposed scheduling algorithms. Our algorithm also produces schedules that have a lower makespan than pure task- and data-parallel schedules. For task graphs with known optimal schedules or lower bounds on the makespan, our algorithm generates schedules that are closer to the optima than other scheduling approaches.

Non-Degeneracy of Pollard Rho Collisions

The Pollard ρ algorithm is a widely used algorithm for solving discrete logarithms on general cyclic groups, including elliptic curves. Recently the first nontrivial runtime estimates were provided for it, culminating in a sharp bound for the collision time on a cyclic group of order n [4, 5]. In this paper, we show that for n satisfying a mild arithmetic condition, the collisions guaranteed by these results are nondegenerate with high probability: that is, the Pollard ρ algorithm successfully finds the discrete logarithm.

A Distributed Fair Scheduling Scheme With a New Analysis Model in IEEE 802.11 Wireless LANs

In this paper, we study the performance of 802.11 wireless local area networks (WLANs) and propose a distributed packet scheduling scheme that improves the short-term fairness among stations in a WLAN. First, we present a completely different analytical model to analyze the performance of 802.11 distributed coordination function (DCF), which models the channel contention dynamics of an IEEE 802.11 system, rather than the individual user perspective adopted by most previous analytical works based on Markov chains. Our model is much simpler than previous Markov chain models that have been used popularly to model the DCF's binary exponential backoff (BEB) procedure. In our model, the backoff stage and the value of the backoff counter are represented as a queueing system and the residual service time at the queue, respectively. Then, we model an entire 802.11 WLAN as a closed queuing network and derive the performance of the 802.11 DCF based on the queueing theory. Next, by exploiting the analytic results, we develop a simple distributed runtime estimation scheme that effectively calculates the number of contending stations. Finally, we propose a distributed packet scheduling scheme that improves the short-term fairness to the IEEE 802.11 network. The scheme utilizes the token bucket mechanism and controls the packet arrival rate sent down to the medium access control (MAC) layer from the logical link control (LLC) layer by adjusting the token-generation rate according to the measured channel contention level. Performance evaluation results show that our scheduling scheme provides a higher degree of short-term fairness as well as throughput improvement.

A probabilistic and adaptive scheduling algorithm using system-generated predictions for inter-grid resource sharing

Rapid advancement and more readily availability of Grid technologies have encouraged many businesses and researchers to establish Virtual Organizations (VO) and make use of their available desktop resources to solve computing intensive problems. These VOs, however, work as disjointed and independent communities with no resource sharing between them. We, in previous work, have proposed a fully decentralized and reconfigurable Inter-Grid framework for resource sharing among such distributed and autonomous Grid systems (Rao et al. in ICCSA, [2006]). The specific problem that underlies in such a collaborating Grids system is scheduling of resources as there is very little knowledge about availability of the resources due to the distributed and autonomous nature of the underlying Grid entities. In this paper, we propose a probabilistic and adaptive scheduling algorithm using system-generated predictions for Inter-Grid resource sharing keeping collaborating Grid systems autonomous and independent. We first use system-generated job runtime estimates without actually submitting jobs to the target Grid system. Then this job execution estimate is used to predict the job scheduling feasibility on the target system. Furthermore, our proposed algorithm adapted itself to the actual resource behavior and performance. Simulation results are presented to discuss the correctness and accuracy of our proposed algorithm.

Fairness enforcement in switch on event multithreading

The need to reduce power and complexity will increase the interest in Switch On Event multithreading (coarse-grained multithreading). Switch On Event multithreading is a low-power and low-complexity mechanism to improve processor throughput by switching threads on execution stalls. Fairness may, however, become a problem in a multithreaded processor. Unless fairness is properly handled, some threads may starve while others consume all of the processor cycles. Heuristics that were devised in order to improve fairness in simultaneous multithreading are not applicable to Switch On Event multithreading. This paper defines the fairness metric using the ratio of the individual threads' speedups and shows how it can be enforced in Switch On Event multithreading. Fairness is controlled by forcing additional thread switch points. These switch points are determined dynamically by runtime estimation of the single threaded performance of each of the individual threads. We analyze the impact of the fairness enforcement mechanism on aggregate IPC and weighted speedup. We present simulation results of the performance of Switch On Event multithreading. Switch On Event multithreading achieves an average aggregate IPC increase of 26% over single thread and 12% weighted speedup when no fairness is enforced. In this case, a sixth of our runs resulted in poor fairness in which one thread ran extremely slowly (10 to 100 times slower than its single-thread performance), while the other thread's performance was hardly affected. By using the proposed mechanism, we can guarantee fairness at different levels of strictness and, in most cases, even improve the weighted speedup.

IMPROVED RUNTIME AND TRANSFER TIME PREDICTION MECHANISMS IN A NETWORK ENABLED SERVERS MIDDLEWARE

In this paper we address the problem of accurately estimating the runtime and communication time of a client request in a Network Enabled Server (NES) middleware such as GridSolve. We use a template based model for the runtime estimation and a client-server communication test for the transfer time estimation. We implement these two mechanisms in GridSolve and test them on a real testbed. Experiments show that they allow for significant improvement in terms of client execution time on various scenarios.