ECS: Energy-Efficient Computing Systems
This research is supported in part by the National Science Foundation (NSF) under Grant CCF-0702781, the New Mexico Institute of Mining and Technology under Grant 103295, by Intel Corporation under Grant 2005-04-070, and by Altera Corporation under an equipment grant.
With the advances of computing technology, the demand on computing systems for high performance and low energy consumption exponentially increases. Next-generation computing systems require innovative energy-efficient scheduling techniques for resource management. In this project, new job and packets scheduling algorithms are designed and implemented for embedded systems, cluster computing platforms, and wireless networks. The algorithms are aimed at achieving the best tradeoffs between energy conservation and high performance for energy-efficient computing systems. The algorithms are evaluated through extensive experiments based on both synthetic benchmark/traces and real-world applications. Our approaches to conserving energy for energy-efficient computing systems include the following:
1. Energy-Efficient Duplication-Based Scheduling for Clusters. [Back to Top]
Optimizing energy consumption has become a major concern in designing economical and environmentally friendly clusters for a wide variety of applications. In recognition that network interconnects are a major part of todays clusters, we propose in this study novel scheduling strategies aimed at reducing energy consumption for the entire interconnect of a cluster. Scheduling precedence constrained parallel tasks on clusters is technically challenging because of, in part, high communication overhead in parallel tasks running on clusters. Therefore, duplication heuristics are applied to schedule parallel tasks to minimize communication overhead. However, most of the available duplication algorithms are developed with consideration of schedule lengths, completely ignoring energy consumption of clusters. In this regard, we design two energy-aware duplication scheduling algorithms, called EADUS and TEBUS, to schedule precedence constrained parallel tasks with a complexity of O(n2), where n is the number of tasks in a parallel task set. Unlike existing duplication-based scheduling algorithms that replicate all the possible predecessors of each task, the proposed algorithms judiciously replicate predecessors of a task if the duplication can help in conserving energy. Our energy-aware scheduling strategies are conducive to balancing the scheduling length and energy consumption of a set of precedence constrained parallel tasks. We conducted extensive experiments using synthetic benchmarks and real-world applications to compare our algorithms with an existing approach. Experimental results based on simulated clusters demonstrate the effectiveness and practicality of the proposed duplication scheduling strategies.
X.-J. Ruan, X. Qin, M. Nijim, Z.-L. Zong, and K. Bellam, Proc. 16th IEEE Int'l Conference on Computer Communications and Networks (ICCCN), Honolulu, Hawaii, Aug. 2007.
Z.-L. Zong, X. Qin*, M. Nijim, X.-J. Ruan, K. Bellam, and M. Alghamdi, Proc. 36th International Conference on Parallel Processing (ICPP), Sept. 2007.
Z.-L. Zong, M. Nijim, and X. Qin, Cluster Computing: The Journal of Networks, Software Tools and Applications. Submitted May 2006; revised Jan. 2007; accepted March 2007.
Z.-L. Zong, A. Manzanares, B. Stinar, and X. Qin, Proc. IEEE 8th International Conference on Cluster Computing (Cluster'06), Sept. 2006.
Z.-L. Zong, M. Nijim, M. Alghamdi, and X. Qin, Proc. 2006 the 7th Symposium of the Los Alamos Computer Science Institute, Santa Fe, NM, Oct. 2006.
2. An Energy-Delay Tunable Task Allocation Strategy for Embedded Systems. [Back to Top]
Applications with energy and low-latency constraints are emerging in various networked embedded systems like digital signal processing, vehicle tracking, and infrastructure monitoring. However, conventional energy-driven task allocation schemes for a cluster of embedded nodes only concentrate on energy-saving when making allocation decisions. Consequently, the length of the schedules could be very long, which is unfavourable or in some situations even not tolerated. In this work, we address the issue of allocating a group of tasks on a heterogeneous embedded system with an objective of energy-saving and short-latency. A novel task allocation strategy, or BEATA (Balanced Energy-Aware Task Allocation), is developed to find an optimal allocation that minimizes overall energy consumption while confining the length of schedule to an ideal range. In addition, we proposed mathematical models to describe a system framework, a set of tasks, energy consumption, and heterogeneity, which are used by BEATA to measure energy dissipation caused by both computation and communication activities. Experimental results show that BEATA significantly improves the performance of embedded systems in terms of energy-saving and schedule length over existing allocation schemes.
T. Xie and X. Qin, IEEE Transactions on Computers. Submitted Nov. 2006; revised April 2007; accepted July 2007.
T. Xie, X. Qin, and M. Nijim, Proc. 35th International Conference on Parallel Processing (ICPP), Columbus, Ohio, Aug. 2006.
3. Energy-Efficient Storage Systems [Back to Top]
Improving the power consumption attributes of embedded and real-time systems has become an important area of interest as processors and other system components have become increasingly powerful and demanding in their energy consumption.
3.1 Power Aware Disk Scheduling Algorithms. [Back to Top]
The focus of this project is on examining the characteristics of real-time hard-disk scheduling algorithms, as hard-disk power usage can amount to a substantial portion of the total system power consumption. Conventional disk scheduling algorithms typically either disregard power consumption completely, or at best apply a fairly naïve power management policy. Given the potential implications for improving the efficiency and longevity of real-world disk systems, we investigate the problem of scheduling a set of independent real-time disk requests such that the total power consumption is minimized, while the efficacy of the disk system is not compromised. We define a power consumption model that can reasonably approximate the performance characteristics of real-world disks. Next, we discuss two power-aware power management policies, I/O Burstiness for Energy Conservation (IBEC) and Speed-Aware Real-time Disk Scheduling for energy conservation (SARDS), which integrate differing power management policies into the disk scheduling algorithms for real-time I/O-intensive applications. Furthermore, to evaluate the performance of the proposed algorithms against existing solutions and ensure that the efficacy of the system is not compromised, we incorporate the earliest deadline first (EDF) and least laxity first (LLF) scheduling policies into SARDS and IBEC to implement power-aware real-time scheduling algorithms. Experimental results from real-world traces and synthetically generated workloads show that the dynamic algorithms have the potential to substantially reduce power consumption over existing scheduling algorithms and power management policies without compromising the overall performance of disk systems.
A. Roth and X. Qin, IEEE Transactions on Computers. Submitted Aug. 2006; revised Feb. 2007.
A. Roth, T. Xie, M. Nijim, and X. Qin, ACM Transactions on Embedded Computing Systems. Submitted Feb. 2006.
3.2 Energy-Efficient Disk Buffer Disks. [Back to Top]
Huge energy consumption has become a critical bottleneck for further applying large-scale cluster systems to build new data centers. Among various components of a data center, storage subsystems are one of the biggest consumers of energy. In this project, we propose a novel buffer-disk based framework for large-scale and energy-efficient parallel storage systems. To validate the efficiency of the proposed framework, a buffer-disk scheduling algorithm is designed and implemented. Our algorithm can provide more opportunities for underlying disk power management schemes to save energy by keeping a large number of idle data disks in sleeping mode as long as possible. The trace-driven simulation results based on a revised disksim simulator show that this new framework can significantly improves the energy efficiency of large-scale parallel storage systems.
Z.-L.Zong, M.E. Briggs, N.W. O'Connor, and X. Qin, Proc. 21st Int'l Parallel and Distributed Processing Symp. (IPDPS), 8th IEEE Int'l Workshop Parallel and Distributed Scientific and Engineering Computing, Long Beach, CA, March 2007.
Z.-L. Zong, M.E. Briggs, N.W. O'Connor, X. Qin, M. Alghamdi, and Y.-M. Yang, the Commodity Cluster Symposium 2007 (CCS2007), Annapolis, Maryland, July 2007.
4. Energy-Efficient Packet Transmissions in Real-Time Wireless Networks [Back to Top]
Reducing energy consumption has become a major goal in designing modern real-time wireless networks. The focus of this study is to investigate the power and real-time issues in wireless networks. The study aims to develop a rich variety of scheduling schemes to reduce energy dissipation while meeting timing constraints of real-time applications in wireless networks. In what follows, we describe our solutions to the energy problem in the context of real-time wireless networks.
Reducing Power Consumption in Real-Time Wireless Networks.
In this work we addressed the issue of scheduling real-time messages in wireless networks subject to timing and power constraints. We developed a novel energy-aware message scheduling scheme, or PARM (Power-aware Real-time Message), which generates optimal schedules minimizing both power consumption and the probability of missing deadlines for real-time messages. In addition, we extended a power consumption model to calculate power consumption rates in accordance to message transmission rates. Experimental results show that PARM significantly improves the performance in terms of missed rate, energy efficiency, and overall performance over four baseline message scheduling schemes.
M. Alghamdi, Z. Zong, K. Bellam, and X. Qin. Submitted.
X. Qin, M. Alghamdi, M. Nijim, Z.-L. Zong, and K. Bellam, Proc. the 26th IEEE Int'l Performance Computing and Communications Conf. (IPCCC'07), New Orleans, Louisiana, April 2007.
X. Qin, M. Alghamdi, T. Xie, M. Nijim, and Z.-L. Zong, IEEE Transactions on Wireless Communications. Submitted Jan. 2006; revised Aug. 2006.
M. Alghamdi, T. Xie, X. Qin, ACM Int'l Symp. Modeling, Analysis and Simulation of Wireless and Mobile Sys. (MSWiM), Workshop Wireless Multimedia Networking and Performance Modeling, Oct. 2005.
Updated on 7/29/2007