Abstract: A power management system schedules the voltage and frequency of processors in a data processing system based on two criteria. The first criterion is a prediction of the performance that the work currently running on the processor will experience at the different frequencies that are available. The second criterion is a system-wide constraint on the total power budget allocated to processors. Based on these criteria, low-level code sets the frequency and voltage of the processors in the system to match what the operating system is currently running on them.

Abstract: A method, computer program product, and a data processing system for optimizing task throughput in a multi-processor system. A performance metric is calculated based on performance counters measuring characteristics of a task executed at one of a plurality of processor frequencies available in the multi-processor system. The characteristics measured by the performance counters indicate activity in the processor as well as memory activity. A performance metric provides a means using measured data at one available frequency to predict performance at another processor frequency available in the multi-processing system. Performance loss minimization is used to assign a particular task to a particular frequency. Additionally, the present invention provides a mechanism for priority load balancing of tasks in a manner that minimizes cumulative performance loss incurred by execution of all tasks in the system.

Abstract: A method of managing power in a data processing system includes monitoring a system parameter indicative of power consumption. Responsive to determining that the parameter differs from a specified threshold, a system guest, such as an operating system, is forced to release a portion of its allocated system memory. The portion of system memory released by the guest is then reclaimed by the system. The reclaimed system memory and the resulting decrease in allocated memory may enable the system to reduce system memory power consumption. The operating system may de-allocate a portion of system memory when a balloon code device driver executing under the operating system requests the operating system to allocate memory to it. The system memory allocated to the balloon device driver is then reclaimed by supervisory code such as a hypervisor.

Abstract: A method and system for machine memory power and availability management in a processing system supporting multiple virtual machines provides a mechanism for supporting memory power management and memory failure management in large scale systems. Upon a decision to take a memory module off-line or place the module in an increased-latency state for power management, or upon a notification that a memory module has failed or been taken off-line or has had latency increased by another power management control mechanism, a hypervisor that supports multiple virtual machines checks the use of pages by each virtual machine and its guest operating system by using a reverse mapping. The hypervisor determines which virtual machines are using a particular machine memory page and may re-map the machine memory page to another available machine page, or may notify the virtual machines that the memory page has become or is becoming unavailable via a fault or other notification mechanism.

Abstract: A method and system for energy management via energy-aware process scheduling provides per-process energy use/power dissipation control to manage system energy requirements and thermal conditions without throttling overall system performance. Use of energy by a particular process is measured or is estimated from resource requirements that are determined by the operating system or reported by the application owning the process. The scheduler then determines whether or not to allocate execution slices to the process in conformity with the measured or estimated energy requirements of the process. The scheduler may insert “idle” execution slices to reduce energy use/power dissipation or may prefer low energy-use processes over high energy-use processes. Pragmatic faults may be issued as warnings from the operating system to an application to indicate that energy requirements need to be curtailed.