Performance aware smart framework to improve CPU power efficiency in static display reading mode
By setting logic components on the PCD's integrated circuit chip, detecting buffer submission and hysteresis statistics in static display reading mode, and performing power saving actions, it solves the problem that the CPU frequently enters and exits low power in static display reading mode by portable computing devices in static display reading mode. The invalid power consumption problem caused by the mode improves CPU power efficiency.
Patent Information
- Application Number
- CN202280094745.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-24
- Publication Date
- 2024-11-26
AI Technical Summary
Portable computing devices (PCDs) are in static display reading mode because the CPUs of readers Apps and Internet web browsers frequently enter and exit the low power mode, resulting in invalid power consumption. The existing technology cannot effectively manage power, resulting in low CPU power efficiency.
Set up logic components on the integrated circuit chip of the PCD. By detecting whether there are active display buffer submissions and cost-saving actions in the static display reading mode, analyze the hysteresis statistical values of the CPU entering and exiting the low-power mode. If the threshold is exceeded, Then perform power cost saving actions, such as limiting the number of CPU cores and disabling LPM mode.
It effectively reduces the CPU power cost of PCD, improves CPU power efficiency, and reduces the invalid power consumption in static display reading mode.
Smart Images

Figure CN119032331A_ABST
Abstract
Description
[0001] Related technical description
[0002] Portable computing devices (PCDs), typically including mobile phones, tablet portable computers (i.e., iPads), and laptop computers, all run on batteries when not being charged. PCD users often use various reader applications (reader apps), reader application package kits (APKs), and Internet web browsers installed on the PCD to read content (i.e., novels, newspapers, documents, etc.). PCDs have emerged in a variety of platforms, and the way they implement power management schemes varies depending on the specific platform of the PCD and the specific reader app or Internet web browser being used by the user to read content at a given time.
[0003] These reader apps, APKs, and internet web browsers operating in reader mode often consume a significant amount of battery power for the PCD. It has been observed that even when the reader app, APK, or internet web browser is in a static display reading mode, the reader app specific threads / processes cause the central processing unit (CPU) processing core cluster (CPU and cluster) to enter and exit a low power mode (LPM) state multiple times per frame. The CPU and cluster enter and exit the LPM state multiple times per frame resulting in significant power consumption.
[0004] Some reader apps or APKs are configured to trigger a power management quality of service (PMQoS) process when the reader app detects a buffer commit, which prevents the CPU and cluster LPM from entering / exiting. However, since there is no buffer commit when the reader app is in static display reading mode, the PMQoS process is not triggered when the reader app or APK is in static display reading mode. As the term is used in this disclosure, buffer commit refers to a series of activities from (App / Ui→Command Buffer→Graphics or Compute Unit or AI Unit→N-Dimensional Rendering Result→Display Buffer / Frame Buffer→Display). When the actual content changes, this is observed on the display. The UI / application needs to submit a "command buffer" to the graphics engine (e.g., GPU, CDSP, etc.) to render the content to generate the required N-dimensional output or graphics content requested by the submitter (e.g., UI / application), which is ultimately passed to the display buffer / frame buffer for final content display. In this context, this final change of the display buffer or frame buffer is called "buffer commit."
[0005] Some other reasons why reader apps and Internet web browsers consume power inefficiently in static display reading mode are: (1) housekeeping and wakeup tasks are distributed across all cores available in the CPU, resulting in utilization of all CPU cores even when there is no gain in performance; and / or (2) housekeeping or wakeup tasks are not well organized or grouped to use the CPU's system resources. The traditional kernel scheduler / framework in PCD is not able to handle the APK-specific patterns of wakeup or housekeeping behavior generated by the APK. The traditional kernel scheduler / framework in Android / Apple operating system (iOS) cannot see the APK / APP behavior in user space and the operating system (OS)-specific workload.
[0006] Therefore, these reader apps, APKs, and internet web browsers in reader mode do not provide the required PMQoS in the PCD when operating in static display reading mode, which results in inefficient power usage in the PCD. Systems and methods are needed for improving CPU power efficiency in a PCD when the PCD is running an app or internet web browser operating in reader mode when the PCD is operating in static display reading mode. Summary of the invention
[0007] Disclosed herein are systems, methods, computer-readable media, and other examples for improving CPU power efficiency in a PCD when the PCD is running an App or an Internet web browser operating in a static display reading mode.
[0008] An exemplary system for reducing central processing unit (CPU) power cost in a PCD may include a logic component disposed on an integrated circuit (IC) chip of the PCD. The logic component may be configured to determine whether the PCD enters a static display reading mode when running a mobile device App or an Internet web browser operating in a reader mode, and in the case where the PCD enters a static display reading mode, determine whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without reducing the visual performance of the display of the PCD, and determine whether there is an active display buffer submission. If it is determined that the display KPI margin indicates that one or more cost saving actions can be performed without reducing visual performance and there is no active display buffer submission, the logic component analyzes the hysteresis statistics associated with the CPU entering and exiting LPM to calculate the total number of LPM entries and exits during a preselected time window. The logic component compares the total number with a first preselected threshold (TH) value, and in the case where the total number exceeds the first preselected TH value, performs one or more power cost saving actions.
[0009] An exemplary embodiment of a method for reducing CPU power costs in a PCD may include, in a logic component disposed on an integrated circuit (IC) chip of the PCD:
[0010] Make sure the PCD is running a mobile device software application (App) or an Internet web browser in reader mode
[0011] Whether to enter the static display reading mode, and in the case where the PCD enters the static display reading mode:
[0012] determining (1) whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without degrading visual performance of a display of the PCD; and (2) whether there is an active display buffer commit;
[0013] If it is determined that the display KPI margin indicates that one or more cost saving actions can be performed without degrading visual performance and there are no active display buffer commits:
[0014] analyzing hysteresis statistics associated with CPU entry into a low power mode (LPM) and exiting the LPM to calculate a total number of LPM entries and exits equal to the number of times the CPU entered the LPM during a preselected time window plus the number of times the CPU exited the LPM during the time window; and
[0015] The total number is compared to a first preselected TH value, and in the event that the total number exceeds the first preselected TH value, one or more power cost saving actions are performed.
[0016] An exemplary embodiment of a non-transitory computer-readable medium includes computer instructions for execution by a logic component disposed on an IC chip of a PCD to reduce CPU power costs in the PCD. The computer instructions may include a first instruction set that determines whether the PCD enters a static display reading mode when running a mobile device App or an Internet web browser operating in a reader mode, and in the event that the PCD enters the static display reading mode, further includes a second instruction set that determines (1) whether the display KPI margin of the PCD indicates that one or more cost-saving actions can be performed without reducing the visual performance of the display of the PCD; and (2) whether there is an active display buffer submission. If the second instruction set is used to determine that the display KPI margin indicates that one or more cost-saving actions can be performed without reducing visual performance and there is no active display buffer submission, the computer instructions may also include a third instruction set that analyzes hysteresis statistics associated with CPU entry and exit of LPM to calculate the total number of LPM entries and exits during a preselected time window. The computer instructions may also include a fourth set of instructions that compares the total number to a first TH value and, if the total number exceeds a first preselected TH value, performs one or more power cost saving actions.
[0017] These and other features and advantages will become apparent from the following description, drawings, and claims. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] In the drawings, like reference numerals refer to like parts throughout the various views unless otherwise indicated. For reference numerals with letter character designations, such as "102A" or "102B," the letter character designations may distinguish between two similar parts or elements in the same figure. The letter character designations of reference numerals may be omitted when it is intended that a reference numeral encompass all parts having the same reference numeral in all figures.
[0019] Figure 1 is a graph having a vertical axis showing the number of times a CPU or CPU core cluster of a PCD running a particular Reader App enters and exits the LPM state within the time window shown on the horizontal axis.
[0020] Figure 2 is a table showing some operational and performance statistics per frame for several reader APKs operating in static display reading mode, as well as the power cost of each APK due to inefficient utilization of CPU cores and due to large numbers of LPM entries and exits.
[0021] Figure 3An example of a PCD, such as a mobile phone or smart phone, is illustrated in which exemplary embodiments of systems, methods, computer-readable media, and other examples of providing a performance-aware intelligence framework for reducing power costs may be provided.
[0022] Figure 4 is a flow chart of a performance-aware intelligence framework algorithm according to a representative implementation scheme.
[0023] Figure 5A and Figure 5B Together they form a flow chart that represents a performance-aware smart framework method for reducing power costs according to an exemplary embodiment. DETAILED DESCRIPTION
[0024] The present disclosure discloses systems, methods, and computer-readable media for reducing CPU power costs in a PCD by determining when the PCD enters a static display reading mode when running an App or an Internet web browser operating in reader mode. When the PCD remains in static display reading mode, the logic of the PCD determines (1) whether the display key performance indicator (KPI) margin of the PCD indicates that one or more cost-saving actions can be performed without reducing the visual performance of the PCD's display; and (2) whether there is an active display buffer submission. If the answers to (1) and (2) are yes and no, respectively, the hysteresis statistics associated with the CPU entering and exiting LPM are analyzed to calculate the total number of CPU LPM entries and exits during a preselected time window. The total number is compared to a first preselected threshold (TH) value, and if the total number exceeds the first preselected TH value, one or more power cost-saving actions are performed.
[0025] In the following specific embodiments, for the purpose of explanation rather than limitation, exemplary or representative embodiments that disclose specific details are described, so as to provide a thorough understanding of the embodiments according to the present teachings. The word "exemplary" is used herein to mean "used as an example, instance, or illustration". The word "exemplary" may be used synonymously with "exemplary" herein. Any aspect described as "exemplary" herein may not be interpreted as being preferred or having advantages over other aspects. However, it will be apparent to those skilled in the art who benefit from the present disclosure that other embodiments that depart from the specific details disclosed herein according to the present teachings are still within the scope of the appended claims. In addition, the description of known devices and methods may be omitted to avoid blurring the description of the example embodiments. Such methods and devices are obviously within the scope of the present teachings.
[0026] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting. The defined terms are in addition to the technical and scientific meanings of the defined terms as commonly understood and accepted in the technical field of the present teachings.
[0027] As used in the specification and the appended claims, the terms "a", "an", and "the" include both singular and plural referents unless the context clearly dictates otherwise. Thus, for example, "a device" includes one device and a plurality of devices.
[0028] Relative terms may be used to describe the relationship of various elements to each other as illustrated in the figures. These relative terms are intended to encompass different orientations of the device and / or elements in addition to the orientation depicted in the figures.
[0029] It will be understood that when an element is described as being “connected to” or “coupled to” or “electrically coupled to” another element, the elements may be directly connected or coupled or intervening elements may be present.
[0030] The term "memory" or "memory device" as used herein is intended to mean a non-transitory computer-readable storage medium capable of storing computer instructions or computer code for execution by one or more processors. References to "memory" or "memory device" herein should be interpreted as one or more memories or multiple memory devices. For example, the memory may be multiple memories within the same computer system. The memory may also be multiple memories distributed in multiple computer systems or computing devices.
[0031] The term "processor" as used herein encompasses an electronic component capable of executing a computer program or executable computer instructions. References herein to a computer comprising a "processor" should be interpreted as one or more processors or processing cores. For example, a processor may be a multi-core processor. A processor may also refer to a collection of processors within a single computer system or distributed across multiple computer systems. The term "computer" should also be interpreted as possibly referring to a collection or network of computers or computing devices, each of which includes one or more processors. The instructions of a computer program may be executed by multiple processors, which may be within the same computer or distributed across multiple computers.
[0032] A computing device may include multiple subsystems, cores, or other components. Such a computing device may be, for example, a portable computing device ("PCD") such as a laptop or palmtop computer, a cellular or smart phone, a portable digital assistant, a portable gaming console (e.g., an extended reality (XR) device, a virtual reality (VR) device, an augmented reality (AR) device, or a mixed reality (MR) device), etc.
[0033] The term "App" used in the remainder of this disclosure refers to a software application designed to run on a PCD, including but not limited to Apps designed to run on Windows, iOS, and Android mobile platforms, and APKs designed to run only on the Android platform.
[0034] Multiple subsystems, cores, or other components of a computing device may be included within the same integrated circuit chip or in different chips. A "system on a chip" or "SoC" is an example of a chip that integrates numerous components to provide system-level functionality. For example, a SoC may include one or more types of processors, such as a central processing unit ("CPU"), a graphics processing unit ("GPU"), a digital signal processor ("DSP"), and a neural processing unit ("NPU"). A SoC may include other processing subsystems, such as a transceiver or "modem" subsystem that provides wireless connectivity.
[0035] The computing device may include resources shared between SoC processors or other processing subsystems. For example, the processors may share access to the main memory or system memory of the computing device. The processors may also be associated with local cache memory.
[0036] Figure 1 1 is a graph 100 having a vertical axis showing the number of times a CPU or CPU core cluster of a PCD running a particular reader App enters and exits the LPM state within a time window shown on the horizontal axis. Figure 1 , any change in direction of the plot 101 corresponds to an LPM state entry or exit. For example, the locations circled by circles 102 to 106 on the plot 101 correspond to LPM state entry or exit. The figure shows that given a single frame duration, such as 120 frames per second (fps) = 8 milliseconds (ms), there are a very large number of changes in the LPM mode. For this particular reader app of this example, there are about 32 changes in direction in the plot 101 during this time window corresponding to about 32 LPM state entries or exits, which results in a relatively large amount of power consumption. Similar results can be obtained for other commonly used reader apps, APKs, and Internet web browsers running on the PCD.
[0037] The following apply to reasons why an app or internet web browser may be running in static display reading mode:
[0038] 1. Traditional kernel schedulers / frameworks cannot handle app-specific wakeup patterns or housekeeping behaviors;
[0039] 2. The traditional kernel scheduler / framework in Android / iOS platforms has no visibility into app behavior in user space and OS-specific workloads, which is expected as there is an incentive to keep the scheduler framework lightweight, but this comes at the cost of power consumption; and
[0040] 3. App-specific system state (e.g., performance state) is unknown to the kernel scheduler / framework.
[0041] In current state-of-the-art PCDs, during static display reading mode operation of a reader app and an internet web browser in reading mode, the CPU power cost will be higher if (1) housekeeping tasks and wakeups are distributed across all cores available in the CPU, resulting in utilization of all CPU cores even if no performance gain is achieved; (2) housekeeping tasks or wakeups are not well organized or grouped to most efficiently use the CPU's system resources; and (3) there are abnormal or random core behaviors (e.g., CPU core behaviors), such as unnecessary entry / exit of LPM mode states, which is critical for any SOC in the power domain.
[0042] For all these reasons, for apps and internet web browsers operating in reading mode, power usage in static display reading mode scenarios is adversely affected. The performance-aware smart framework algorithms and logic components of the present disclosure provide a solution to this problem. It should be noted that although representative or exemplary embodiments are described herein with reference to reader apps, the inventive principles and concepts are applicable to and can be beneficially used with any app that can sometimes enter a static display reading mode and any internet web browser that operates in a reading mode and can sometimes enter a static display reading mode.
[0043] Figure 2200 shows some operations and performance statistics per frame for four different commercially available reader APKs operating in static display reading mode. The first commercially available APK reader (Reader 1 in Table 200) utilizes eight CPU cores, enters / exits the LPM state mode approximately thirty times, and consumes 32 milliwatts (mW) of CPU power per frame. The second commercially available reader APK (Reader 2 in Table 200) utilizes seven CPU cores, enters / exits the LPM state approximately twenty-eight times, and consumes 28 mW of CPU power per frame. The third commercially available reader APK (Reader 3 in Table 200) utilizes eight CPU cores, enters / exits the LPM state approximately thirty-two times, and consumes 32 mW of CPU power per frame. The fourth commercially available reader APK (Reader 4 in Table 200) utilizes seven CPU cores, enters / exits the LPM state approximately twenty-seven times, and consumes 28 mW of CPU power per frame.
[0044] Therefore, from Figure 1 and Figure 2 From the above description, it can be seen that the static display reading mode scenario requires a power management scheme to provide more efficient CPU core utilization and reduce the number of LPM state entries / exits that occur, thereby reducing CPU power costs.
[0045] According to a representative embodiment of the present disclosure, the performance-aware smart framework algorithm is initiated or triggered when the PCD enters a static display reading mode while running an App or when a whitelisted App is launched. In the context of the present disclosure, the term "whitelisted" means that the App has been designated by the original equipment manufacturer (OEM) or user of the PCD as an App that (1) can benefit from running the performance-aware smart framework algorithm both at all times when the App is running and during certain usage scenarios of the App, and (2) during these times, power cost-saving actions taken by the performance-aware smart framework algorithm (discussed below) will not degrade visual performance. The "whitelisted" name is stored somewhere in the memory of the PCD that is accessible to the processor of the PCD, which can trigger the performance-aware smart framework algorithm and cause it to jump directly to taking one or more power cost-saving actions in the absence of active buffer submission.
[0046] When the performance-aware smart framework algorithm is initiated or triggered, it determines whether (1) there is sufficient display key performance index (KPI) margin to perform one or more power cost saving actions; and (2) whether there are any current display buffers committed. If inquiries (1) and (2) are answered yes and no, respectively, the algorithm analyzes the LPM state entry / exit statistics (i.e., for Figure 1The hysteresis of the data of the type of drawing shown) within a time window (e.g., each frame) and determining whether to perform one or more power cost saving actions based on the hysteresis analysis. A variety of power cost saving actions can be performed, as shown below. Figure 5A and Figure 5B Let's describe a few examples.
[0047] As noted above, if the App is whitelisted, this is an indication that a determination has previously been made that the answer to (1) is yes. In this case, the determination of (1) can be equivalent to a determination that the App is whitelisted, which is determined when the App is launched. This allows the Performance Aware Smart Framework algorithm to jump immediately to the determination of (2) when the algorithm is triggered. In this case, if the answer to the determination made at (2) is no, the algorithm analyzes the hysteresis of the LPM state entry / exit statistics within a time window (e.g., each frame) and decides whether to perform one or more power cost saving actions based on the hysteresis analysis.
[0048] According to a representative embodiment, hysteresis of CPU LPM entry / exit can be used in the following manner: counting the number of CPU LPM entries and exits that occurred in the most recent frame, comparing the number to a preselected threshold (TH) value, and if the number exceeds the preselected TH value, performing at least one of the following cost saving actions: (1) limiting the number of CPU cores that can be utilized when operating in a static display reading mode; (2) selecting one or more of the CPU cores for utilization among the limited number of CPU cores that can be utilized; and (3) limiting the number of LPM state entries / exits that may occur in at least one of the CPU cores that have been selected for utilization, for example, by disabling LPM mode for one or more of the cores being utilized.
[0049] The following will refer to Figure 5A and Figure 5B As described in more detail, performing one or more of these power cost saving actions when the PCD is in a static display reading mode significantly improves CPU power utilization. As will be discussed in more detail below, one or more additional power cost saving actions may also be performed to further reduce CPU power costs during the static display reading mode.
[0050] The disclosed systems and methods for providing a performance-aware smart framework to reduce power costs may be implemented in a PCD, although the inventive principles and concepts are not limited in the type of device implementing the method. Figure 3An example of a PCD 300 is illustrated, such as a mobile phone, a smart phone, or a portable game console (such as an extended reality (XR) device, a virtual reality (VR) device, an augmented reality (AR) device, or a mixed reality (MR) device), in which exemplary embodiments of systems, methods, computer-readable media, and other examples providing a performance-aware intelligence framework for reducing power costs may be implemented. For clarity, Figure 3 Some interconnects, signals, etc. are not shown. Although PCD 300 is shown as an example, other embodiments of systems, methods, computer-readable media, and other examples of using a performance-aware intelligence framework to reduce power costs may be provided in other types of computing devices or systems.
[0051] PCD 300 may include SoC 302. SoC 302 may include CPU 304, NPU 305, GPU 306, DSP 307, analog signal processor 308, modem / modem subsystem 354, or other processors. CPU 304 may include one or more CPU cores, such as first CPU core 304A, second CPU core 304B, etc. to Nth CPU core 304N, where N is a positive integer greater than or equal to 1.
[0052] Cores 304A to 304N may be configured to perform certain operations according to the performance-aware smart framework of the present disclosure. According to a representative embodiment, SoC 302 includes a performance-aware smart framework logic component 310 that communicates with CPU 304 to execute a performance-aware smart framework algorithm. CPU cores 304A to 304N perform other operations of the type that they would normally perform in a PCD. Alternatively or in addition, any of the processors (such as NPU 305, GPU 306, DSP 307, etc.) may perform some or all of these operations. Performance-aware smart framework logic component 310 may also be implemented within CPU 304 rather than in a separate logic block.
[0053] The PCD 300 may include a display controller 309 and a touch screen controller 312 coupled to the CPU 304. A touch screen display 314 external to the SoC 302 may be coupled to the display controller 310 and the touch screen controller 312. The PCD 300 may also include a video decoder 316 coupled to the CPU 304. A video amplifier 318 may be coupled to the video decoder 316 and the touch screen display 314. A video port 320 may be coupled to the video amplifier 318. A universal serial bus (“USB”) controller 322 may also be coupled to the CPU 304, and a USB port 324 may be coupled to the USB controller 322. A subscriber identity module (“SIM”) card 326 may also be coupled to the CPU 304.
[0054] One or more memories 328 may be coupled to CPU 304. One or more memories 304 may include both volatile memory and non-volatile memory. Examples of volatile memory include static random access memory ("SRAM") and dynamic random access memory ("DRAM"). Such memory may be external to SoC 302 or internal to SoC 302. One or more memories 328 may include local cache memory or system-level cache memory.
[0055] A stereo audio CODEC 334 may be coupled to the analog signal processor 308. Further, an audio amplifier 336 may be coupled to the stereo audio CODEC 334. A first stereo speaker 338 and a second stereo speaker 340 may be coupled to the audio amplifier 336, respectively. In addition, a microphone amplifier 342 may be coupled to the stereo audio CODEC 334, and a microphone 344 may be coupled to the microphone amplifier 342. A frequency modulation ("FM") radio tuner 346 may be coupled to the stereo audio CODEC 334. An FM antenna 348 may be coupled to the FM radio tuner 346. Further, a stereo headset 350 may be coupled to the stereo audio CODEC 334. Other devices that may be coupled to the CPU 304 include one or more digital (e.g., CCD or CMOS) cameras 352.
[0056] A modem or RF transceiver 354 may be coupled to the analog signal processor 308 and the CPU 304. An RF switch 356 may be coupled to the RF transceiver 354 and the RF antenna 358. Additionally, a keypad 360, a mono headset with a microphone 362, and a vibrator device 364 may be coupled to the analog signal processor 308. The SoC 302 may have one or more internal or on-chip thermal sensors 370. A power source 374 and a PMIC 376 may supply power to the SoC 302.
[0057] The firmware or software may be stored in any of the above memories, or may be stored in local memory directly accessible to the processor hardware on which the software or firmware is executed. The execution of such firmware or software may control aspects of any of the above methods or configure aspects of any of the above systems. Any such memory or other non-transitory storage medium having firmware or software stored therein in a computer-readable form for execution by processor hardware may be an example of a "computer-readable medium" as that term is understood in the patent dictionary.
[0058] Figure 44 is a flow chart of a performance-aware smart framework algorithm according to a representative embodiment. According to a representative embodiment of the present disclosure, when the PCD enters a static display reading mode, or when a whitelisted App is launched, at block 401, the performance-aware smart framework algorithm is triggered or started. In the step represented by block 402, when the PCD remains in the static display reading mode, the performance-aware smart framework logic component 310 ( Figure 3 ) answers the following inquiries: (1) Is there sufficient display KPI headroom available for logic component 310 to perform one or more cost saving actions of the present disclosure without causing one or more display KPIs to drop to unacceptable levels indicating reduced visual performance? (2) Is there an active display buffer submission currently? As noted above, inquiry (1) can be equivalent to a previous determination at launch time that an App is whitelisted.
[0059] At block 403, if queries (1) and (2) of block 402 are answered yes and no, respectively, then logic component 310 analyzes the CPU LPM entry / exit hysteresis statistics to determine whether the number of CPU LPM entries and exits exceeds a preselected TH value, and if so, causes one or more power cost saving actions to be performed, as will be described below with reference to Figure 5A and Figure 5B Detailed description
[0060] If the PCD exits the static display reading mode at any time during the process represented by blocks 401 through 403, the algorithm executed by logic component 310 resets and returns to block 401 to continue monitoring for the PCD to enter the static display reading mode. Figure 4 The process depicted in can be performed on a per-frame basis or once for every preselected number of frames.
[0061] As noted above, a variety of power cost saving actions may be performed, as will be described below with reference to Figure 5A and Figure 5B Typically, the cost saving actions include one or more of the following: (1) limiting the number of CPU cores that can be utilized when operating in a static display reading mode; and (2) limiting the number of LPM state entries / exits that may occur in at least one of the CPU cores that have been selected for utilization, for example, by disabling LPM mode for one or more of the cores being utilized. Figure 5A and Figure 5B Other actions discussed may also be performed, as will be discussed in more detail below.
[0062] Figure 5A is a flow chart showing the logic component 310 ( Figure 3 ) is a performance-aware smart framework method for reducing power costs. At the step represented by block 501, a determination is made as to whether the PCD is in a static display reading mode. This can be performed by Figure 4 402 in the static display reading mode, although other methods may be used to determine when the PCD is in static display reading mode. In the case where the PCD is in static display reading mode, the process proceeds to box 502, where App housekeeping tasks and wakeups are classified into three buckets, namely, (1) system-specific / OS-specific task buckets; (2) display performance buckets and (3) App-specific buckets. App or APK wakeups or housekeeping tasks primarily fall into these three buckets.
[0063] Bucket (1) corresponds to OS / system specific tasks or App / APK wake-up / housekeeping tasks. These are system specific housekeeping tasks or wake-ups that are generic in nature, for example, specific to the Android or iOS platform or specific to the OS of the PCD. Bucket (2) corresponds to display performance tasks or wake-up / housekeeping tasks. More specifically, these are performance tasks related to the display thread that is responsible for affecting fps / stuttering or otherwise affecting the visual performance of the display. Bucket (3) corresponds to APK specific tasks or wake-up / housekeeping tasks. These tasks involve additional framework specific housekeeping tasks or wake-ups associated with App / APK and system tasks that are executed when more than one App or APK is running on the PCD.
[0064] Categorizing these tasks and wakeups into these three buckets is fairly straightforward and well understood by those skilled in the art, but the classification process will be briefly described. For bucket (2), these tasks and wakeups can be identified and whitelisted for a given OS. For bucket (3), when an App or APK is launched, the process ID (PID) associated with its peers can be extracted and maintained in a list. For bucket (1), these tasks and wakeups typically correspond to all remaining tasks / threads that do not belong to bucket (1) or bucket (2), i.e., they belong to the OS.
[0065] At the step represented by box 503, the display thread is monitored on a per-frame basis for key performance indicators (KPIs) for housekeeping tasks and wakeups that have been classified into bucket (2). At box 504, a determination is made as to whether the analysis made at box 503 indicates that there is sufficient headroom to perform power cost saving actions without degrading visual display performance, such as by determining whether the analysis indicates that no fps drops / stutters or similar issues have occurred that adversely affect visually perceptible performance. The manner in which the amount of headroom can be determined and analyzed to determine whether the KPI is met is understood by those skilled in the art. An example of a manner for clicking on headroom in an Android PCD is to use the following command: SFen-deq wrt Vsync boundary for given panel.
[0066] While the process is being performed at box 504, a determination is made at box 505 as to whether the analysis performed at box 503 indicates an active buffer commit. If boxes 504 and 505 are answered with "yes" and "no", respectively, this is an indication that: (1) the display performance KPI indicates that the display performance is of sufficient quality that performing one or more power cost saving actions will not adversely affect visual performance; and (2) the PCD is still in static display reading mode. If boxes 504 and 505 are answered with "no" and "yes", respectively, this is an indication that: (1) performing one or more power cost saving actions may adversely affect visual performance; and (2) the PCD is no longer in static display reading mode. In the latter case, the performance-aware smart framework algorithm may exit or return to box 501, at which time the performance-aware smart framework logic component 310 ( Figure 3 ) continues to monitor the PCD to determine when it re-enters static display reading mode.
[0067] If the answers to blocks 504 and 505 are "yes" and "no," respectively, the process proceeds to block 506. At block 506, the hysteresis of the CPU LPM entry and exit statistics is analyzed on a per-frame basis to determine whether the number of CPU LPM entries and exits exceeds a preselected TH value. Figure 4 As indicated in block 403 of FIG. 1 , in the event that the number of CPU LPM entries and exits exceeds a preselected TH value, one or more power cost saving actions are performed, as described below with reference to FIG. Figure 5B As described in blocks 507 to 510.
[0068] It should be noted that in Figure 5A and Figure 5BWhen the performance-aware smart framework algorithm represented by the first run is executed, the determination represented by box 504 can be equivalent to the determination made when an App that has been previously designated as a whitelisted App is launched. In this case, the step represented by box 504 can be skipped when the performance-aware smart framework algorithm is executed for the first time for the whitelisted App.
[0069] Examples of specific power cost saving actions that can be taken are given by Figure 5B The flowchart shown in . Figure 5B is a flow chart of a CPU core power optimization algorithm that is executed according to an exemplary embodiment to perform power cost saving actions in the event that the number of CPU LPM entries and exits exceeds a preselected TH value, assuming sufficient KPI margin and no active display buffer commits.
[0070] In by Figure 5B At the step represented by box 507 in , in the case where the number of CPU LPM entries and exits exceeds the preselected TH value, the minimum and maximum number of CPU cores to be used when the PCD is in static display reading mode are set. These minimum and maximum numbers are referred to as MIN-MAX number pairs hereinafter because they consist of number pairs. The MIN-MAX number pairs can be pre-generated by determining the minimum number of CPU cores that need to be active and the maximum number that may need to be active for an App or an Internet web browser in reading mode so that they can operate without visually perceptible performance degradation. It should be noted that this type of approach is not possible for a general scheduler in a conventional PCD system because the conventional scheduler does not understand the behavior of the App and can be controlled by the App, which distributes tasks across all CPU cores without any intelligence. An example of a MIN-MAX number pair according to the present disclosure is (2,3), which means that two cores are always enabled, and a third core can be enabled if the scheduler sleep state requires any spike in workload.
[0071] At the step represented by box 508, one or more CPU cores are selected and enabled according to the MIN-MAX number pair set at box 507. Once the MIN-MAX number pair has been set at box 507, the CPU core can be selected randomly at box 508 or based on one or more other considerations or criteria. For example, the CPU cores in the PCD system can be sorted from lower to higher according to their respective power costs, in which case a lower-ranked CPU core can be selected instead of a higher-ranked CPU core. For another example, the CPU cores can be sorted from low to high according to the current workload, demand and / or utilization reported by the underlying scheduler / framework, in which case a lower-ranked CPU core can be selected instead of a higher-ranked CPU core. A combination of these two situations is also possible, and other considerations or criteria can be considered when selecting and enabling CPU cores.
[0072] The steps represented by blocks 507 and 508 are examples of power cost saving actions that may be performed. An additional power cost saving action that may be performed is to selectively disable LPM in one or more of the CPU cores selected and enabled at block 508. At the step represented by block 509, Figure 5A The number of LPM entries and exits determined at block 506 is compared to a second TH value, and LPM for one or more of the selected cores is disabled to prevent those cores from entering and exiting LPM during the static display reading mode. Based on the number of LPM entries / exits that result in a specific power payback for each CPU core, a CPU core power cost per frame may be estimated for each CPU core. Then, at block 509, that number may be used as a second TH value, such that if the number counted at block 506 exceeds the second TH value, LPM is disabled in the corresponding core. Block 510 represents a step of disabling the LPM state in one or more of the cores based on the process executed in block 509.
[0073] The second TH value may be based on a pre-estimate that is configured into the logic 310, for example, in a factory setup. The estimation and setting of the second TH value may also be performed on the fly during operation of the PCD using internal logic of the PCD or using logic external to the PCD. The first TH value and the second TH value may be stored in a memory of the PCD, such as in the memory 328 ( Figure 3 ) or resides in another memory device on or accessible by the PCD 300.
[0074] It should be noted that additional CPU power cost saving actions can be performed, such as selectively disabling LPM in cache memories used by the cores, dynamically ramping down the CPU operating frequency for actual workloads, dynamically reducing bus bandwidth (BW), dynamically reducing BW voting for DDR and all types of cache memories, dynamically disabling LPM state and ramping down operating frequency for any IP in the PCD (such as, for example, GPU, NSP, CDSP, ADSP, modem, WLAN, sensor and / or DPU) and / or dynamically increasing the number of enabled CPU cores and / or the amount of cache memory based on the current workload and resource utilization.
[0075] Specific implementation examples are described in the following numbered clauses:
[0076] 1. A system for reducing central processing unit (CPU) power cost in a portable computing device (PCD), the system comprising:
[0077] A logic component disposed on an integrated circuit (IC) chip of the PCD, the logic component being configured to:
[0078] determining whether the PCD enters a static display reading mode when running a mobile device software application (App) or an Internet web browser operating in a reader mode, and in the event that the PCD enters the static display reading mode:
[0079] determining (1) whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without degrading visual performance of a display of the PCD; and (2) whether there is an active display buffer commit;
[0080] If it is determined that the display KPI margin indicates that one or more cost saving actions can be performed without degrading visual performance and there are no active display buffer commits:
[0081] analyzing hysteresis statistics associated with the CPU entering a low power mode (LPM) and exiting the LPM to calculate a total number of LPM entries and exits, the total number being equal to the number of times the CPU entered the LPM during a preselected time window plus the number of times the CPU exited the LPM during the time window; and
[0082] The total number is compared to a first preselected threshold (TH) value, and in the event that the total number exceeds the first preselected TH value, one or more power cost saving actions are performed.
[0083] 2. The system of clause 1, wherein performing the one or more power cost saving actions comprises at least one of:
[0084] limiting the total number of cores of the CPU permitted to be utilized by the App or Internet web browser operating in reader mode to a limited number of CPU cores;
[0085] Among the limited number of CPU cores, enabling one or more CPU cores among the CPU cores, so that the one or more enabled CPU cores can be used by the App or Internet web browser running in the reader mode; and
[0086] Limits the total number of times at least one enabled CPU core is allowed to enter and exit LPM.
[0087] 3. The system of clause 2, wherein the step of limiting the total number of times at least one enabled CPU core is allowed to enter and exit LPM is performed by disabling LPM in the at least one enabled CPU core.
[0088] 4. A system according to clause 2 or 3, wherein the step of limiting the total number of cores of the CPU allowed to be utilized by the App or Internet web browser operating in reader mode is performed by:
[0089] Setting the minimum number of CPU cores required for the App or Internet web browser to operate in a static display reading mode without reducing performance to a level that can be visually perceived by a user of the PCD; and
[0090] The maximum number of CPU cores that may be required for the App or Internet web browser to operate in static display reading mode is set to ensure that performance will not be reduced to a level that can be visually perceived by the user of the PCD.
[0091] 5. The system of clauses 2 to 4, wherein the step of enabling one or more of the CPU cores is performed by:
[0092] Among the limited number of CPU cores, selecting the minimum number of CPU cores to be enabled based on one or more criteria; and
[0093] Enable the selected minimum number of CPU cores.
[0094] 6. The system of clause 5, wherein the one or more criteria include ranking each of the limited number of CPU cores according to power cost, such that the minimum number of CPU cores to be enabled is selected based on the corresponding ranking.
[0095] 7. A system according to clause 5, wherein the one or more criteria include ranking each of the limited number of CPU cores according to at least one of current workload, current demand, and current utilization reported by a scheduler of the App or Internet web browser.
[0096] 8. The system of any one of clauses 4 to 7, wherein the step of disabling LPM in the at least one enabled CPU core comprises:
[0097] comparing the total number of LPM entries and exits to a second TH value;
[0098] selecting one or more enabled CPU cores for which LPM is to be disabled based on said comparison of said total number of LPM entries and exits and said second TH value; and
[0099] Disable LPM in the selected one or more enabled CPU cores.
[0100] 9. The system of clause 8, wherein the second TH value is set based on an estimate, for each CPU core, of a respective power payoff associated with LPM entry and exit of the respective CPU core.
[0101] 10. A method for reducing central processing unit (CPU) power costs in a portable computing device (PCD), the method comprising, in a logic component disposed on an integrated circuit (IC) chip of the PCD:
[0102] determining whether the PCD enters a static display reading mode when running a mobile device software application (App) or an Internet web browser operating in a reader mode, and in the event that the PCD enters the static display reading mode:
[0103] determining (1) whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without degrading visual performance of a display of the PCD; and (2) whether there is an active display buffer commit;
[0104] If it is determined that the display KPI margin indicates that one or more cost saving actions can be performed without degrading visual performance and there are no active display buffer commits:
[0105] analyzing hysteresis statistics associated with the CPU entering a low power mode (LPM) and exiting the LPM to calculate a total number of LPM entries and exits, the total number being equal to the number of times the CPU entered the LPM during a preselected time window plus the number of times the CPU exited the LPM during the time window; and
[0106] The total number is compared to a first preselected threshold (TH) value, and in the event that the total number exceeds the first preselected TH value, one or more power cost saving actions are performed.
[0107] 11. The method of clause 10, wherein performing the one or more power cost saving actions comprises at least one of:
[0108] limiting the total number of cores of the CPU permitted to be utilized by the App or Internet web browser operating in reader mode to a limited number of CPU cores;
[0109] Among the limited number of CPU cores, enabling one or more CPU cores among the CPU cores, so that the one or more enabled CPU cores can be used by the App or Internet web browser running in the reader mode; and
[0110] Limits the total number of times at least one enabled CPU core is allowed to enter and exit LPM.
[0111] 12. The method of clause 11, wherein the step of limiting the total number of times at least one enabled CPU core is allowed to enter and exit LPM comprises:
[0112] The LPM is disabled in the at least one enabled CPU core.
[0113] 13. The method of any one of clauses 11 or 12, wherein the step of limiting the total number of cores of the CPU allowed to be utilized by the App or Internet web browser operating in reader mode comprises:
[0114] Setting the minimum number of CPU cores required for the App or Internet web browser to operate in a static display reading mode without reducing performance to a level that can be visually perceived by a user of the PCD; and
[0115] The maximum number of CPU cores that may be required for the App or Internet web browser to operate in static display reading mode is set to ensure that performance will not be reduced to a level that can be visually perceived by the user of the PCD.
[0116] 14. The method of any one of clauses 11 to 13, wherein the step of enabling one or more of the CPU cores comprises:
[0117] Among the limited number of CPU cores, selecting the minimum number of CPU cores to be enabled based on one or more criteria; and
[0118] Enable the selected minimum number of CPU cores.
[0119] 15. The method of clause 14, wherein the one or more criteria include a ranking of each of the limited number of CPU cores according to power cost, such that the minimum number of CPU cores to be enabled is selected based on the respective rankings.
[0120] 16. A method according to clause 14, wherein the one or more criteria include ranking each of the limited number of CPU cores according to at least one of current workload, current demand, and current utilization reported by a scheduler of the App or Internet web browser.
[0121] 17. The method of any one of clauses 12 to 16, wherein the step of disabling LPM in the at least one enabled CPU core comprises:
[0122] comparing the total number of LPM entries and exits to a second TH value;
[0123] selecting one or more enabled CPU cores for which LPM is to be disabled based on said comparison of said total number of LPM entries and exits and said second TH value; and
[0124] Disable LPM in the selected one or more enabled CPU cores.
[0125] 18. The method of clause 17, wherein the second TH value is set based on an estimate, for each CPU core, of a respective power penalty associated with LPM entry and exit of the respective CPU core.
[0126] 19. A non-transitory computer readable medium comprising computer instructions for execution by a logic component disposed on an integrated circuit (IC) chip of a portable computing device (PCD) to reduce a central processing unit (CPU) power cost in the PCD, the computer instructions comprising:
[0127] A first instruction set, the first instruction set determines whether the PCD enters a static display reading mode when running a mobile device software application (App) or an Internet web browser operating in a reader mode, and in the event that the PCD enters the static display reading mode:
[0128] determining, using a second set of instructions, (1) whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without degrading visual performance of a display of the PCD; and (2) whether there is an active display buffer commit;
[0129] If it is determined using the second instruction set that the display KPI margin indicates that one or more cost saving actions can be performed without degrading visual performance and there is no active display buffer commit:
[0130] analyzing, using a third instruction set, hysteresis statistics associated with the CPU entering a low power mode (LPM) and exiting the LPM to calculate a total number of LPM entries and exits, the total number being equal to a number of times the CPU entered the LPM during a preselected time window plus a number of times the CPU exited the LPM during the time window; and
[0131] The total number is compared to a first preselected threshold (TH) value using a fourth set of instructions, and in the event that the total number exceeds the first preselected TH value, one or more power cost saving actions are performed.
[0132] 20. The non-transitory computer-readable medium of clause 19, wherein to perform the one or more power cost saving actions, the fourth set of instructions includes at least one of:
[0133] a fifth instruction set that limits the total number of cores of the CPU permitted to be utilized by the App or Internet web browser operating in reader mode to a limited number of CPU cores;
[0134] a sixth instruction set, the sixth instruction set enabling one or more CPU cores of the limited number of CPU cores;
[0135] and
[0136] A seventh instruction set is provided to disable LPM in at least one of the enabled CPU cores.
[0137] 21. The non-transitory computer-readable medium of clause 20, wherein the fifth set of instructions limits the total number of cores of the CPU permitted to be utilized by the App or Internet web browser operating in reader mode by:
[0138] Setting the minimum number of CPU cores required for the App or Internet web browser to operate in a static display reading mode without reducing performance to a level that can be visually perceived by a user of the PCD; and
[0139] The maximum number of CPU cores that may be required for the App or Internet web browser to operate in static display reading mode is set to ensure that performance will not be reduced to a level that can be visually perceived by the user of the PCD.
[0140] 22. The non-transitory computer-readable medium of any one of clauses 20 to 21, wherein the sixth set of instructions enables one or more of the CPU cores by:
[0141] Among the limited number of CPU cores, selecting the minimum number of CPU cores to be enabled based on one or more criteria; and
[0142] Enable the selected minimum number of CPU cores.
[0143] 23. A non-transitory computer-readable medium according to clause 22, wherein the one or more criteria include ranking each of the limited number of CPU cores according to power cost, thereby selecting the minimum number of CPU cores to be enabled based on the corresponding ranking.
[0144] 24. A non-transitory computer-readable medium as described in clause 22, wherein the one or more criteria include ranking each of the limited number of CPU cores according to at least one of current workload, current demand, and current utilization reported by a scheduler of the App or Internet web browser.
[0145] 25. The non-transitory computer-readable medium of any one of clauses 20 to 24, wherein the seventh set of instructions disables LPM in the at least one enabled CPU core by:
[0146] comparing the total number of LPM entries and exits to a second TH value;
[0147] selecting one or more enabled CPU cores for which LPM is to be disabled based on said comparison of said total number of LPM entries and exits and said second TH value; and
[0148] Disable LPM in the selected one or more enabled CPU cores.
[0149] 26. The non-transitory computer-readable medium of clause 25, wherein the second TH value is set based on an estimate, for each CPU core, of a respective power payoff associated with LPM entry and exit of the respective CPU core.
[0150] It should be noted that the principles and concepts of the present invention have been described with reference to representative embodiments, but the principles and concepts of the present invention are not limited to the representative embodiments described herein. Although the principles and concepts of the present invention have been illustrated and described in detail in the drawings and the foregoing description, such illustrations and descriptions should be considered illustrative or exemplary rather than restrictive; the present invention is not limited to the disclosed embodiments. By studying the drawings, the disclosure and the appended claims, those skilled in the art will be able to understand and implement other variations of the disclosed embodiments.
Claims
1. A system for reducing central processing unit (CPU) power cost in a portable computing device (PCD), the system comprising: A logic component disposed on an integrated circuit (IC) chip of the PCD, the logic component being configured to: determining whether the PCD enters a static display reading mode when running a mobile device software application (App) or an Internet web browser operating in a reader mode, and in the event that the PCD enters the static display reading mode: determining (1) whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without degrading visual performance of a display of the PCD; and (2) whether there is an active display buffer commit; If it is determined that the display KPI margin indicates that one or more cost saving actions can be performed without degrading visual performance and there are no active display buffer commits: analyzing hysteresis statistics associated with the CPU entering a low power mode (LPM) and exiting the LPM to calculate a total number of LPM entries and exits, the total number being equal to a number of times the CPU entered the LPM during a preselected time window plus a number of times the CPU exited the LPM during the time window; as well as The total number is compared to a first preselected threshold (TH) value, and in the event that the total number exceeds the first preselected TH value, one or more power cost saving actions are performed.
2. The system of claim 1 , wherein performing the one or more power cost saving actions comprises at least one of: limiting the total number of cores of the CPU permitted to be utilized by the App or Internet web browser operating in reader mode to a limited number of CPU cores; Among the limited number of CPU cores, enabling one or more CPU cores among the CPU cores, so that the one or more enabled CPU cores can be used by the App or Internet web browser running in the reader mode; and Limits the total number of times at least one enabled CPU core is allowed to enter and exit LPM.
3. The system of claim 2, wherein the step of limiting the total number of times at least one enabled CPU core is allowed to enter and exit LPM is performed by disabling LPM in the at least one enabled CPU core.
4. The system of claim 3, wherein the step of limiting the total number of cores of the CPU that are allowed to be utilized by the App or Internet web browser operating in reader mode is performed by: Setting the minimum number of CPU cores required for the App or Internet web browser to operate in a static display reading mode without reducing performance to a level that can be visually perceived by a user of the PCD; and The maximum number of CPU cores that may be required for the App or Internet web browser to operate in static display reading mode is set to ensure that performance will not be reduced to a level that can be visually perceived by the user of the PCD.
5. The system of claim 4, wherein the step of enabling one or more of the CPU cores is performed by: Among the limited number of CPU cores, selecting the minimum number of CPU cores to be enabled based on one or more criteria; and Enable the selected minimum number of CPU cores. 6 . The system of claim 5 , wherein the one or more criteria include a ranking of each of the limited number of CPU cores according to power cost, such that the minimum number of CPU cores to be enabled is selected based on the corresponding ranking.
7. The system of claim 5, wherein the one or more criteria include ranking each of the limited number of CPU cores according to at least one of a current workload, a current demand, and a current utilization reported by a scheduler of the App or Internet web browser.
8. The system of claim 4, wherein the step of disabling LPM in the at least one enabled CPU core comprises: comparing the total number of LPM entries and exits to a second TH value; selecting one or more enabled CPU cores for which LPM is to be disabled based on said comparison of said total number of LPM entries and exits and said second TH value; as well as Disable LPM in the selected one or more enabled CPU cores.
9. The system of claim 8, wherein the second TH value is set based on an estimate, for each CPU core, of a respective power penalty associated with LPM entry and exit of the respective CPU core.
10. A method for reducing central processing unit (CPU) power costs in a portable computing device (PCD), the method comprising, in a logic component disposed on an integrated circuit (IC) chip of the PCD: determining whether the PCD enters a static display reading mode when running a mobile device software application (App) or an Internet web browser operating in a reader mode, and in the event that the PCD enters the static display reading mode: determining (1) whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without degrading visual performance of a display of the PCD; and (2) whether there is an active display buffer commit; If it is determined that the display KPI margin indicates that one or more cost saving actions can be performed without degrading visual performance and there are no active display buffer commits: analyzing hysteresis statistics associated with the CPU entering a low power mode (LPM) and exiting the LPM to calculate a total number of LPM entries and exits, the total number being equal to a number of times the CPU entered the LPM during a preselected time window plus a number of times the CPU exited the LPM during the time window; as well as The total number is compared to a first preselected threshold (TH) value, and in the event that the total number exceeds the first preselected TH value, one or more power cost saving actions are performed.
11. The method of claim 10, wherein performing the one or more power cost saving actions comprises at least one of: limiting the total number of cores of the CPU permitted to be utilized by the App or Internet web browser operating in reader mode to a limited number of CPU cores; Among the limited number of CPU cores, enabling one or more CPU cores among the CPU cores, so that the one or more enabled CPU cores can be used by the App or Internet web browser running in the reader mode; and Limits the total number of times at least one enabled CPU core is allowed to enter and exit LPM.
12. The method of claim 11, wherein the step of limiting the total number of times at least one enabled CPU core is allowed to enter and exit LPM comprises: The LPM is disabled in the at least one enabled CPU core.
13. The method of claim 12, wherein the step of limiting the total number of cores of the CPU allowed to be utilized by the App or Internet web browser operating in reader mode comprises: Setting the minimum number of CPU cores required for the App or Internet web browser to operate in a static display reading mode without reducing performance to a level that can be visually perceived by a user of the PCD; as well as The maximum number of CPU cores that may be required for the App or Internet web browser to operate in static display reading mode is set to ensure that performance will not be reduced to a level that can be visually perceived by the user of the PCD.
14. The method of claim 13, wherein the step of enabling one or more of the CPU cores comprises: Among the limited number of CPU cores, selecting the minimum number of CPU cores to be enabled based on one or more criteria; as well as Enable the selected minimum number of CPU cores. 15 . The method of claim 14 , wherein the one or more criteria include a ranking of each of the limited number of CPU cores according to power cost, such that the minimum number of CPU cores to be enabled is selected based on the respective rankings.
16. The method of claim 14, wherein the one or more criteria include ranking each of the limited number of CPU cores according to at least one of a current workload, a current demand, and a current utilization reported by a scheduler of the App or Internet web browser.
17. The method of claim 13, wherein the step of disabling LPM in the at least one enabled CPU core comprises: comparing the total number of LPM entries and exits to a second TH value; selecting one or more enabled CPU cores for which LPM is to be disabled based on said comparison of said total number of LPM entries and exits and said second TH value; as well as Disable LPM in the selected one or more enabled CPU cores.
18. The method of claim 17, wherein the second TH value is set based on an estimate, for each CPU core, of a respective power penalty associated with LPM entry and exit of the respective CPU core.
19. A non-transitory computer readable medium comprising computer instructions for execution by a logic component disposed on an integrated circuit (IC) chip of a portable computing device (PCD) to reduce a central processing unit (CPU) power cost in the PCD, the computer instructions comprising: A first instruction set, the first instruction set determines whether the PCD enters a static display reading mode when running a mobile device software application (App) or an Internet web browser operating in a reader mode, and in the event that the PCD enters the static display reading mode: determining, using a second set of instructions, (1) whether a display key performance indicator (KPI) margin of the PCD indicates that one or more cost saving actions can be performed without degrading visual performance of a display of the PCD; and (2) whether there is active display buffer submission; If it is determined using the second instruction set that the display KPI margin indicates that one or more cost saving actions can be performed without degrading visual performance and there is no active display buffer commit: analyzing, using a third instruction set, hysteresis statistics associated with the CPU entering a low power mode (LPM) and exiting the LPM to calculate a total number of LPM entries and exits, the total number being equal to a number of times the CPU entered the LPM during a preselected time window plus a number of times the CPU exited the LPM during the time window; as well as The total number is compared to a first preselected threshold (TH) value using a fourth set of instructions, and in the event that the total number exceeds the first preselected TH value, one or more power cost saving actions are performed.
20. The non-transitory computer readable medium of claim 19, wherein to perform the one or more power cost saving actions, the fourth set of instructions includes at least one of: a fifth instruction set that limits the total number of cores of the CPU permitted to be utilized by the App or Internet web browser operating in reader mode to a limited number of CPU cores; a sixth instruction set, the sixth instruction set enabling one or more CPU cores of the limited number of CPU cores; and A seventh instruction set is provided to disable LPM in at least one of the enabled CPU cores.
21. The non-transitory computer readable medium of claim 20, wherein the fifth set of instructions limits the total number of cores of the CPU that are allowed to be utilized by the App or Internet web browser operating in reader mode by: Setting the minimum number of CPU cores required for the App or Internet web browser to operate in a static display reading mode without reducing performance to a level that can be visually perceived by a user of the PCD; and The maximum number of CPU cores that may be required for the App or Internet web browser to operate in static display reading mode is set to ensure that performance will not be reduced to a level that can be visually perceived by the user of the PCD.
22. The non-transitory computer readable medium of claim 21, wherein the sixth instruction set enables one or more of the CPU cores by: Among the limited number of CPU cores, selecting the minimum number of CPU cores to be enabled based on one or more criteria; and Enable the selected minimum number of CPU cores.
23. The non-transitory computer-readable medium of claim 22, wherein the one or more criteria include a ranking of each of the limited number of CPU cores according to power cost, such that the minimum number of CPU cores to be enabled is selected based on the corresponding ranking.
24. The non-transitory computer-readable medium of claim 22, wherein the one or more criteria include ranking each of the limited number of CPU cores according to at least one of a current workload, a current demand, and a current utilization reported by a scheduler of the App or Internet web browser.
25. The non-transitory computer readable medium of claim 22, wherein the seventh set of instructions disables LPM in the at least one enabled CPU core by: comparing the total number of LPM entries and exits to a second TH value; selecting one or more enabled CPU cores for which LPM is to be disabled based on said comparison of said total number of LPM entries and exits and said second TH value; and Disable LPM in the selected one or more enabled CPU cores.
26. The non-transitory computer readable medium of claim 25, wherein the second TH value is set based on an estimate, for each CPU core, of a respective power payoff associated with LPM entry and exit of the respective CPU core.
