Abstract: Techniques for monitoring platform energy consumption. One or more operational states of the platform are monitored during a period of time. For each of the one or more operational states, a portion of the period of time that the platform was in the respective one or more operational states is determined. Energy consumption information corresponding to the one or more operational states is retrieved. The energy consumption information and the portions of the period of time are utilized to determine an energy consumption for the period of time.

Abstract: A method, device, and system are disclosed. In one embodiment the method includes grouping multiple memory requests into multiple of memory rank queues. Each rank queue contains the memory requests that target addresses within the corresponding memory rank. The method also schedules a minimum burst number of memory requests within one of the memory rank queues to be serviced when the burst number has been reached in the one of the plurality of memory rank queues. Finally, if a memory request exceeds an aging threshold, then that memory request will be serviced.

Abstract: A new audio playback architecture may be used, which allows the use of much larger buffering than that used by a typical audio subsystem in a computing system to improve power efficiency of the system and at the same time allows to maintain the quality (e.g., fidelity and responsiveness) of the audio playback. The audio controller in the new architecture may be made to report back to the host system a more accurate indication of which audio frame is being set to the audio codec than a currently available audio controller does. Additionally, the controller is capable of re-fetching previously buffered (but not yet transmitted) data. Furthermore, buffers in both the audio controller and the main memory may be dynamically adjusted during playback of audio data and/or for different applications.

Abstract: A method, device, and system are disclosed. In one embodiment the method includes grouping multiple memory requests into multiple of memory rank queues. Each rank queue contains the memory requests that target addresses within the corresponding memory rank. The method also schedules a minimum burst number of memory requests within one of the memory rank queues to be serviced when the burst number has been reached in the one of the plurality of memory rank queues.

Abstract: The memory content may be cached in the non-volatile cache when a computing system is entering S4 state. The non-volatile cache may be coupled to a bus that connects the disk drive with the disk controller. When resuming from S4 state, the memory content may be read from the non-volatile cache rather than from the slow disk drive, which facilitates instant-on resuming for the system. The caching process may be performed in an OS-transparent manner. During the caching process, data with contiguous addresses may be merged into a block of data. A mapping table may be created and stored in the non-volatile cache which includes multiple entries, each for a block of data. The mapping table facilitates data reading from the non-volatile cache to provide instant-on resuming from S4 state.

Abstract: In one embodiment a method for accounting processing resources expended on an activity is disclosed. The method can include determining a task to be performed by a domain, where the task can utilize at least one hardware resource and at least one software resource. The method can monitor and correlate events that are only visible as hardware events with events that are only visible as software events. In one embodiment, this capability is applied to virtual machine configurations on platform power-managed systems to provided correlated platform performance state characteristics on virtual machine, workload or thread level. The method can also combine an output metric of the hardware monitor with an output metric of the software monitor to provide an accounting of resources utilized by the task.

Abstract: A new audio playback architecture may be used, which allows the use of much larger buffering than that used by a typical audio subsystem in a computing system to improve power efficiency of the system and at the same time allows to maintain the quality (e.g., fidelity and responsiveness) of the audio playback. The audio controller in the new architecture may be made to report back to the host system a more accurate indication of which audio frame is being set to the audio codec than a currently available audio controller does. Additionally, the controller is capable of re-fetching previously buffered (but not yet transmitted) data. Furthermore, buffers in both the audio controller and the main memory may be dynamically adjusted during playback of audio data and/or for different applications.

Abstract: A method for managing power data includes determining an amount of power used for a system running an application over a first time period from an operating system. An amount of power used for the system in a baseline state over a second time period is determined from the operating system. A net power consumption of the application is determined from the amount of power used for the system running the application and the amount of power used for the system in the baseline state.

Abstract: A computing system may conserve more power by entering S4 state than S3 state over long periods of inactivity and also have an instant-on capability when assuming from S4 state by using a fast accessible non-volatile cache (e.g., flash memory). Rather than storing memory content to a disk drive, the memory content may be cached in the non-volatile cache when the system is entering S4 state. The non-volatile cache may be coupled to a bus that connects the disk drive with the disk controller. When resuming from S4 state, the memory content may be read from the non-volatile cache rather than from the slow disk drive. Both the caching and resuming processes may be performed in an OS-transparent manner. A mapping table may be created and stored in the non-volatile cache during the caching process to provide efficient reading from the non-volatile cache during the resuming process.

Abstract: The memory content may be cached in the non-volatile cache when a computing system is entering S4 state. The non-volatile cache may be coupled to a bus that connects the disk drive with the disk controller. When resuming from S4 state, the memory content may be read from the non-volatile cache rather than from the slow disk drive, which facilitates instant-on resuming for the system. The caching process may be performed in an OS-transparent manner. During the caching process, data with contiguous addresses may be merged into a block of data. A mapping table may be created and stored in the non-volatile cache which includes multiple entries, each for a block of data. The mapping table facilitates data reading from the non-volatile cache to provide instant-on resuming from S4 state.