Lorch J.R., Smith A.J., Scheduling techniques for reducing processor energy use in MacOS, Wireless Networks, No. 3, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....speed causes a linear reduction in MIPS and energy consumption, the MIPJ metric remains una ected. In systems in which performance degradation is acceptable, power optimizations may be evaluated by showing the system s slowdown, when the optimization is applied, as a function of the power savings [37]. Such a function is useful for computing easily the possible power savings for a speci c acceptable decrease in performance. Finally, the percentage of battery consumption while performing a speci c task can be used in evaluating power optimization on energy constrained systems such as ....

....higher power savings. As an example, the standard scheduling techniques of the operating system may have to be changed in order to maximize the use of the low power processor mode. Lorch and Smith suggested modi cations at the scheduler level in order to improve the power eciency of the MacOS [36, 37]. The Macintosh systems are not able to use the supported power modes eciently for two main reasons. First, MacOS sometimes schedules processes that have no work to do (the events they are waiting for have not arrived yet) Second, the programmers often assume that their application is the ....

[Article contains additional citation context not shown here]

Lorch, J. R., and Smith, A. J. Scheduling techniques for reducing processor energy use in MacOS. Wireless Networks 3, 5 (Oct. 1997), 311{ 324.

....has been a continuous demand for new scheduling policies to address specific requirements of modern OSes and emerging applications. Examples include policies for multimedia and real time applications [2, 3, 6, 10, 14] and energybased policies so as to increase the mission time of portable devices [8, 12, 11]. While the need for new scheduling policies is well recognized, their implementation within an existing OS kernel raises many problems. Based on an analysis of several OS kernels such as RT Linux, Linux and BSD, we have identified the following difficulties in integrating a new scheduling ....

J. Lorch and A. Smith. Scheduling techniques for reducing processor energy use in MacOS. Wireless Networks, 3(5):311--324, October 1997.

....processes within the operating system has been used in prior research [12] in the context of powering down peripherals mainly, and this is the first investigation of such issues for DRAM powermode control. Previous OS based power management studies also considered voltage and frequency scaling [1, 4, 5, 11, 15, 17]. DRAM mode control by the operating system at re scheduling points attempt optimization at a much coarser granularity than a compiler or a pure hardware based approach, where transitions can potentially be effected at memory access granularities. However, it is our belief (and our experimental ....

J. R. Lorch and A. J. Smith. Scheduling techniques for reducing processor energy use in MacOS. Wireless Networks, 3(5):311--324, 1997.

....Delft University of Technology, The Netherlands. E mail: pouwelse,koen,sips ubicom.tudelft.nl. J. Pouwelse is supported by the Dutch Organization for Applied Scientific Research (TNO) Physics and Electronics Laboratory. nificantly reduced. Power savings up to 66 have been reported [2]. A further refinement is to make continuous trade offs between performance and cost. Performance metrics are application dependent, but often a combination of response time and quality is used. Video decoding is used as a case study throughout this paper, and typical quality metrics are ....

....EPS to find the optimal schedule would increase. A heuristic such as EPS will fail to find the optimal schedule in complex workloads with many tasks. For example, when modifying case 2 slightly by changing task B to start at time 2, the insertion of task C will not raise the water above interval [0,2] as it could when realizing that B in turn should push task A aside. Fortunately, such workloads are not common for wearable devices where users typically run one or two concurrent applications. The complexity of the heuristic depends on the number of iterations needed to schedule . In the ....

J. Lorch and A. Smith, "Scheduling techniques for reducing processor energy use in MacOS," Wireless Networks, 1997.

....power management has not been well studied for OS based power management (OSPM) 3] Intuitively, scheduling for power management is to make idle periods clustered and long, instead of scattered and short, so that power management is applicable. Previous scheduling techniques focus on processors [8] [10] 14] or real time systems [4] 12] These algorithms deal with only one service provider the processor; it is unclear how to extend them for multiple devices. The authors do not explain how to integrate the algorithms into existing systems. Furthermore, they unrealistically assume perfect ....

J. R. Lorch and A. J. Smith. Scheduling Techniques for Reducing Processor Energy Use in MacOS. Wireless Networks, 3(5):311--324, 1997.

....the latter schedules computation to increase operand reuse. In [24] the authors propose scheduling for pipelined systems; they derive the conditions when scheduling with buffer insertion reduces power. Several studies investigate the relationship between scheduling and dynamic voltage scaling [25] [27] In [28] the authors discuss the effect of voltage scaling and quality of service. 3) Reducing Memory Power: In addition to IO devices and processors, memory also consumes significant amount of power. Reducing memory power can be achieved by selecting different power performance modes at ....

J. R. Lorch and A. J. Smith, "Scheduling techniques for reducing processor energy use in MacOS," Wireless Networks, vol. 3, no. 5, pp. 311--324, 1997.

....used and appropriately transition them to a lower power consuming mode. Even the software the operating system (OS) compiler and the application has an important role to play in power efficient systems design. The operating system, which plays the role of hardware manager, can schedule jobs [23], allocate and manage memory [19] and control peripherals [20] to reduce overall system power. The compiler can generate code and data transformations to increase idleness of hardware components so that they can be transitioned to low power modes more effectively [7] Finally, algorithmic ....

J. R. Lorch and A. J. Smith. Scheduling Techniques for Reducing Processor Energy Use in MacOS. Wireless Networks, 3(5):311--324, October 1997.

....Scheduling Policies 3 1 Motivation Emerging computing models and applications are continuously challenging the operating system scheduler. Multimedia applications require predictable performance and stringent timing guarantees [12, 16, 19, 36] Embedded systems need to minimize power consumption [15, 29, 35]. Network routers demand isolated execution of active network programs [24] Multiprocessor applications call for processor affinity allocation [33] or performance driven allocation [7] Meeting all these requirements requires specialized scheduling policies, which traditional scheduling ....

J. Lorch and A. Smith. Scheduling techniques for reducing processor energy use in MacOS. Wireless Networks, 3(5):311--324, October 1997.

....of prolonged inactivity. Several methods of extending battery lifetime within the operating system and middleware layer are discussed in (Tiwari et al. 1994, Chandrakasan and Brodersen, 1995, Mehta et al. 1997) Other techniques studied include power aware CPU scheduling (Weiser et al. 1994, Lorch and Smith, 1997) and page allocation (Lebeck et al. 2000) Within the application layer, the power conserving mechanisms tend to be application speci c such as database access (Imielinski et al. 1994, Alonso and Ganguly, 1993) and video processing (Chandrakasan and Brodersen, 1995, Gordon et al. 1996, ....

....simulations show that improvement of 6 to 55 in the Energy Delay metric are obtained using power aware page allocation mechanisms that operate in conjunction with hardware policies. CPU scheduling techniques that attempt to minimize power consumption are presented in (Weiser et al. 1994, Lorch and Smith, 1997). The impact of software architecture on power consumption is studied in (Tiwari et al. 1994, Mehta et al. 1997) 8.2. Application Layer The application layer in a wireless system is responsible for such things as partitioning of tasks between the xed and mobile hosts, audio and video source ....

Lorch, J. and A. J. Smith: 1997, `Scheduling Techniques for Reducing Processor Energy Use in MacOS'. ACM/Baltzer Wireless Networks 3(5), 311-324.

....sound, etc. Re activation of hardware can take some time, which a ects performance (e.g. response time) Using simple power downwhen idle techniques the processor s power consumption can be signi cantly reduced. Depending on the usage pattern, the power savings can amount to a 66 reduction [13]. A re nement is to make continuous trade o s between performance and cost. The user demand (performance) must be supplied at the lowest cost (power consumption) Performance can be expressed as the response time for interactive applications, and as spatial temporal resolution, color depth, and ....

....generated by (bursty) applications, and must derive the optimal settings from external observations, for example, by monitoring the system load and estimating the future demand. This is a non trivial task, for example, for the Mac OS it is hard to determine when no useful computation is occurring [13]. Predicting the future workload from the current situation is dicult, and mispredictions can seriously reduce the gains of voltage scaling as observed in several simulation studies [6, 11, 17, 23] These simulations use an interval based scheduler with a time window of 5 to 100 ms. When idle ....

J. Lorch and A. Smith. Scheduling techniques for reducing processor energy use in MacOS. Wireless Networks, 1997.

....(OSPM) 3] Intuitively, scheduling for power management is to make idle periods clustered and long, instead of scattered and short, so that power management is applicable. Previous International Workshop on Hardware Software Codesign 2000, p 39 43 1 scheduling techniques focus on processors [8] [10] 14] or real time systems [4] 12] These algorithms deal with only one service provider the processor; it is unclear how to extend them for multiple devices. The authors do not explain how to integrate the algorithms into existing systems. Furthermore, they unrealistically assume perfect ....

J. R. Lorch and A. J. Smith. Scheduling Techniques for Reducing Processor Energy Use in MacOS. Wireless Networks, 3(5):311--324, 1997.

....the use of power modes being provided in emerging PADRAM class memory devices. A novel aspect of our work is the cooperative hardware OS approach to exploit PADRAM features. Previous OS level studies focusing on power management include work on scheduling for low power processor modes [31, 32, 46], spindown policies for disks and alternatives [1, 7, 8, 9, 16, 25, 30, 47] and managing wireless communication [18, 24, 42] A consortium of companies has developed a specification [20] that addresses the lower level OS device interface, providing one model for gross system wide power states and ....

J. Lorch and A. J. Smith. Scheduling Techniques for Reducing Processor Energy Use in MacOS. Wireless Networks, 3(5):311--324, October 1997.

....sound, etc. Re activation of hardware can take some time, which affects performance (e.g. response times) Using simple power down when idle techniques the processor s power consumption can be significantly reduced. Depending on the usage pattern, the power savings can amount to a 66 reduction [12]. A refinement is to make continuous tradeoffs between performance and cost. The user demand (performance) must be supplied at the lowest cost (power consumption) Performance can be expressed as the response time for interactive applications, and as spatial temporal resolution, color depth, and ....

....generated by (bursty) applications, and must derive the optimal settings from external observations, for example, by monitoring the system load and estimating the future demand. This is a non trivial task; for the Mac OS it is already hard to determine when no useful computation is occurring [12]. Predicting the future workload from the current situation is difficult, and errors can seriously reduce the gains of voltage scaling as observed in several simulation studies [2, 3, 4, 5] These simulations use an interval based scheduler with a time window of 5 to 100 ms. When idle time is ....

J.R. Lorch, A.J. Smith, "Scheduling techniques for reducing processor energy use in MacOS", Wireless Networks, No. 3, 1997.

....the system to select another decoder. 4 Related work Power optimizations at the OS level is an area of active research. A detailed power breakdown of mobile computers is described in [9] Most research uses the power down when idle principle for a single component such as the disk [5] CPU [8, 11], and wireless link [14] We studied trade o s between two components (decoding and wireless link) Odyssey [3, 10] is an excellent example research project aiming at lowering the power consumption by involving the application. Power reductions are obtained by adapting the application performance ....

Lorch J.R., Smith A.J., Scheduling techniques for reducing processor energy use in MacOS, Wireless Networks, No. 3, 1997.

....multiple power states. We also investigate met Operating System Functionality Energy Efficient Techniques Disk scheduling Spindown policies [18, 6, 5, 14, 11] Security Adaptive cryptographic policy based on computation communication overhead CPU scheduling Voltage scaling, idle power modes [32, 19, 22] Application OS Interaction Agile content negotiation trading fidelity for power, APIs [8] Memory allocation Adaptive placement of memory blocks, switching of hardware energy conservation modes Resource Protection Allocation Fair distribution of battery life among both local and distributed ....

J. Lorch and A. J. Smith. Scheduling Techniques for Reducing Processor Energy Use in MacOS. Wireless Networks, 3(5):311--324, October 1997.

....From a broader perspective, there has been a substantial body of work on hardware power management for mobile computers. This includes efforts that span multiple components [10, 16] as well as those focusing on a single component such as the network [11, 20] disk [4, 5, 6, 12, 25] and CPU [14, 24]. As shown in this paper, such efforts aimed at hardware components are complementary to reducing energy usage through application driven fidelity reduction. 7 Conclusion Relentless pressure to make mobile computers lighter and more compact places severe restrictions on battery capacity. At the ....

Lorch, J. R. and Smith, A. J. Scheduling techniques for reducing processor energy use in MacOS. Wireless Networks, 3(5):311--324, October 1997.

No context found.

Lorch J.R., Smith A.J., Scheduling techniques for reducing processor energy use in MacOS, Wireless Networks, No. 3, 1997.

No context found.

J. Lorch and A. Smith, "Scheduling techniques for reducing processor energy use in MacOS," Wireless Networks, 1997.

No context found.

Lorch, J. R. and Smith, A. J. (1997) Scheduling techniques for reducing processor energy use in MacOS. Wireless Networks, 3, 311--324.

No context found.

J. R. Lorch and A. J. Smith, "Scheduling techniques for reducing processor energy use in MacOS", Wireless Networks,, vol. 3, pp. 311--324, 1997.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC