Results 11  20
of
46
Poweraware replica placement and update strategies in tree networks
, 2011
"... ..."
(Show Context)
How to Schedule When You Have to Buy Your Energy
 In: Proc. of the 13th/14th Workshop on Approximation Algorithms for Comb. Optimization Problems/Randomization and Computation (APPROX/RANDOM
, 2010
"... Abstract. We consider a situation where jobs arrive over time at a data center, consisting of identical speedscalable processors. For each job, the scheduler knows how much income is lost as a function of how long the job is delayed. The scheduler also knows the fixed cost of a unit of energy. The ..."
Abstract

Cited by 8 (1 self)
 Add to MetaCart
(Show Context)
Abstract. We consider a situation where jobs arrive over time at a data center, consisting of identical speedscalable processors. For each job, the scheduler knows how much income is lost as a function of how long the job is delayed. The scheduler also knows the fixed cost of a unit of energy. The online scheduler determines which jobs to run on which processors, and at what speed to run the processors. The scheduler's objective is to maximize profit, which is the income obtained from jobs minus the energy costs. We give a (1+ )speed O(1)competitive algorithm, and show that resource augmentation is necessary to achieve O(1)competitiveness.
Nonclairvoyant speed scaling for batched parallel jobs on multiprocessors
 In CF
, 2009
"... Energy consumption and heat dissipation have become key considerations for modern high performance computer systems. In this paper, we focus on nonclairvoyant speed scaling to minimize flow time plus energy for batched parallel jobs on multiprocessors. We consider a common scenario where the to ..."
Abstract

Cited by 6 (5 self)
 Add to MetaCart
(Show Context)
Energy consumption and heat dissipation have become key considerations for modern high performance computer systems. In this paper, we focus on nonclairvoyant speed scaling to minimize flow time plus energy for batched parallel jobs on multiprocessors. We consider a common scenario where the total power consumption cannot exceed a given budget and the power consumed on each processor is sα when running at speed s. Extending the Equi processor allocation policy, we propose two algorithms: UEqui and NEqui, which use respectively a uniformspeed and a nonuniform speed scaling function for the allocated processors. Using competitive analysis, we show that UEqui is O(P (α−1)/α 2)competitive for flow time plus energy, and NEqui is O ( α lnP)competitive for the same metric when given sufficient power, where P is the total number of processors. Our simulation results confirm that UEqui and NEqui achieve better performance than a straightforward fixedspeed Equi strategy. Moreover, moderate power constraint does not significantly affect the performance of our algorithms.
Algorithms for Energy Saving
, 2010
"... Energy has become a scarce and expensive resource. There is a growing awareness in society that energy saving is a critical issue. This paper surveys algorithmic solutions to reduce energy consumption in computing environments. We focus on the system and device level. More specifically, we study po ..."
Abstract

Cited by 5 (0 self)
 Add to MetaCart
(Show Context)
Energy has become a scarce and expensive resource. There is a growing awareness in society that energy saving is a critical issue. This paper surveys algorithmic solutions to reduce energy consumption in computing environments. We focus on the system and device level. More specifically, we study powerdown mechanisms as well as dynamic speed scaling techniques in modern microprocessors.
Online Deadline Scheduling with Bounded Energy Efficiency
"... Abstract. Existing work on scheduling with energy concern has focused on minimizing the energy for completing all jobs or achieving maximum throughput [19, 2, 7, 13, 14]. That is, energy usage is a secondary concern when compared to throughput and the schedules targeted may be very poor in energy ef ..."
Abstract

Cited by 4 (0 self)
 Add to MetaCart
(Show Context)
Abstract. Existing work on scheduling with energy concern has focused on minimizing the energy for completing all jobs or achieving maximum throughput [19, 2, 7, 13, 14]. That is, energy usage is a secondary concern when compared to throughput and the schedules targeted may be very poor in energy efficiency. In this paper, we attempt to put energy efficiency as the primary concern and study how to maximize throughput subject to a userdefined threshold of energy efficiency. We first show that all deterministic online algorithms have a competitive ratio at least ∆, where ∆ is the maxmin ratio of job size. Nevertheless, allowing the online algorithm to have a slightly poorer energy efficiency leads to constant (i.e., independent of ∆) competitive online algorithm. On the other hand, using randomization, we can reduce the competitive ratio to O(log ∆) without relaxing the efficiency threshold. Finally we consider a special case where no jobs are “demanding ” and give a deterministic online algorithm with constant competitive ratio for this case. 1
Automated techniques for energy efficient scheduling on homogeneous and heterogeneous chip multiprocessor architectures
 in ASPDAC ’08: Proceedings of the 2008 Asia and South Pacific Design Automation Conference. Los Alamitos
"... Abstract — We address performance maximization of independent task sets under energy constraint on chip multiprocessor (CMP) architectures that support multiple voltage/frequency operating states for each core. We prove that the problem is strongly NPhard. We propose polynomial time 2approximat ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
(Show Context)
Abstract — We address performance maximization of independent task sets under energy constraint on chip multiprocessor (CMP) architectures that support multiple voltage/frequency operating states for each core. We prove that the problem is strongly NPhard. We propose polynomial time 2approximation algorithms for homogeneous and heterogeneous CMPs. To the best of our knowledge, our techniques offer the tightest bounds for energy constrained design on CMP architectures. Experimental results demonstrate that our techniques are effective and efficient under various workloads on several CMP architectures. I.
Nonmigratory multiprocessor scheduling for response time and energy
 IEEE Transactions on Parallel and Distributed Systems
, 2008
"... Abstract—Energy usage has been an important concern in recent research on online job scheduling, where processors are allowed to vary the speed dynamically so as to save energy whenever possible. Providing good quality of service such as response time (flowtime) and conserving energy are conflicting ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
(Show Context)
Abstract—Energy usage has been an important concern in recent research on online job scheduling, where processors are allowed to vary the speed dynamically so as to save energy whenever possible. Providing good quality of service such as response time (flowtime) and conserving energy are conflicting objectives. An interesting problem for scheduling is how to optimize an economic tradeoff of flowtime and energy. To this end, the past two years have witnessed significant progress in the singleprocessor setting, and online algorithms with performance close to optimal have been obtained. In this paper, we extend the study of optimizing the tradeoff between flowtime and energy to the multiprocessor setting. We devise and analyze a simple nonmigratory online algorithm that makes use of the classified roundrobin (CRR) strategy to dispatch jobs. Even in the worst case, its performance is within Oðlog PÞ times of the optimal migratory offline algorithm, where P is the ratio of the maximum job size to the minimum job size. Technically speaking, this online result stems from a nontrivial solution to an offline problem of eliminating migration, which is also interesting by itself. Index Terms—Analysis of algorithms and problem complexity, sequencing and scheduling, online computation, energyaware systems. Ç 1
Stochastic analysis of poweraware scheduling
 in Proc. of Allerton
, 2008
"... Abstract—Energy consumption in a computer system can be reduced by dynamic speed scaling, which adapts the processing speed to the current load. This paper studies the optimal way to adjust speed to balance mean response time and mean energy consumption, when jobs arrive as a Poisson process and pro ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
(Show Context)
Abstract—Energy consumption in a computer system can be reduced by dynamic speed scaling, which adapts the processing speed to the current load. This paper studies the optimal way to adjust speed to balance mean response time and mean energy consumption, when jobs arrive as a Poisson process and processor sharing scheduling is used. Both bounds and asymptotics for the optimal speeds are provided. Interestingly, a simple scheme that halts when the system is idle and uses a static rate while the system is busy provides nearly the same performance as the optimal dynamic speed scaling. However, dynamic speed scaling which allocates a higher speed when more jobs are present significantly improves robustness to bursty traffic and misestimation of workload parameters. I.
Scheduling Algorithms for Procrastinators
, 2006
"... This paper presents scheduling algorithms for procrastinators, where the speed that a procrastinator executes a job increases as the due date approaches. We give optimal offline scheduling policies for linearly increasing speed functions. We then explain the computational/numerical issues involved ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
This paper presents scheduling algorithms for procrastinators, where the speed that a procrastinator executes a job increases as the due date approaches. We give optimal offline scheduling policies for linearly increasing speed functions. We then explain the computational/numerical issues involved in implementing this policy. We next explore the online setting, showing that there exist adversaries that force any online scheduling policy to miss due dates. This impossibility result motivates the problem of minimizing the maximum interval stretch of any job; the interval stretch of a job is the job’s flow time divided by the job’s due date minus release time. We show that several common scheduling strategies, including the “hitthehighestnail ” strategy beloved by procrastinators, have arbitrarily large maximum interval stretch. Then we give the “thrashing ” scheduling policy and show that it is a Θ(1) approximation algorithm for the maximum interval stretch.