Dynamic power down management in computing devices

By measuring the current consumption, bandwidth usage and efficiency of the computing device subsystem, and dynamically adjusting the power down count value, the problem of poor performance and power consumption during power down and power on the subsystem in power management in computing devices is solved, and the improvement is achieved Efficient power management.

CN119096217APending Publication Date: 2024-12-06QUALCOMM INC
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202380035723.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2022-05-05
Filing Date
2023-04-03
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

In power management, existing computing devices are difficult to effectively balance power consumption and performance, especially during power down and power-up of memory subsystems, resulting in potential performance impact and poor power consumption.

Method used

By measuring the current consumption, bandwidth usage and efficiency of the subsystem, the idle power count value (IPDC) is determined, and the power outage time of the subsystem is determined based on these measurement values. The polynomial equation system is used to dynamically adjust the power outage value to achieve More efficient power management.

Benefits of technology

More precise power management is achieved, reducing the delay of the memory subsystem, improving performance and power efficiency, and balancing power consumption and performance requirements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119096217A_ABST
    Figure CN119096217A_ABST
Patent Text Reader

Abstract

In a computing device, whether to power down a subsystem may be determined based on how long the subsystem has been idle and based on one or more measurements of subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency. An idle power down count value may be determined based on the measurements using a predetermined relationship between subsystem current draw, subsystem bandwidth usage, and subsystem efficiency. The subsystem may be powered down based on a determination of whether the subsystem has been idle for an interval not less than the idle power down count value.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] Related technical description

[0002] 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 smartphone, a personal digital assistant, a navigation device, a smartbook, a portable game console, a satellite phone, etc.

[0003] 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, a memory subsystem, and the like.

[0004] When a PCD is powered by a battery, power management is an important consideration. "Power management" refers to techniques used to balance power consumption and performance in a PCD and to manage the generation of thermal energy (i.e., heat). One power management technique involves selecting a power mode. For example, in a low power mode, one or more subsystems may be powered off. A memory subsystem is an example of a subsystem that may be powered off when it is determined that lower power consumption has a higher priority than higher performance and powered on when it is determined that higher performance has a higher priority than lower power consumption. The determination of when to power off a memory subsystem may be based on a determination of the activity level (or conversely, the idle level) of the memory subsystem.

[0005] A counter or timer may be used to determine the activity level of the memory subsystem. If a memory transaction has not been initiated within a predetermined number of clock cycles or an "idle power-down count," the memory subsystem may be powered down. Memory transactions occurring while the memory subsystem is powered down may trigger a power-up of the memory subsystem.

[0006] Powering down and powering up the memory subsystem incurs a certain amount of latency, potentially adversely affecting performance. A higher (more conservative) idle power down count results in higher performance at the expense of higher power consumption, while a lower (more aggressive) idle power down count results in lower power consumption at the expense of lower performance. It is desirable to more effectively balance power consumption and performance. Summary of the invention

[0007] Systems, methods, computer-readable media, and other examples are disclosed for dynamic power-off management in a computing device.

[0008] An exemplary method for controlling powering off of a subsystem of a computing device may include determining a subsystem current consumption measurement, a subsystem bandwidth usage measurement, and a subsystem efficiency measurement. The exemplary method may also include determining an idle power-off count ("IPDC") value using a predetermined relationship between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement. The exemplary method may also include powering off the subsystem in response to a determination that the subsystem has been idle for an interval not less than a determined IPDC value.

[0009] An exemplary system for controlling powering off of a subsystem of a computing device may include a processing system and a power controller. The processing system may be configured to determine a subsystem current consumption measurement, a subsystem bandwidth usage measurement, and a subsystem efficiency measurement. The processing system may be further configured to determine an IPDC value using a predetermined relationship between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement. The power controller may be configured to power off the subsystem in response to a determination of whether the subsystem is idle for an interval not less than a determined IPDC value.

[0010] Another exemplary system for controlling powering off of a subsystem of a computing device may include: means for determining a subsystem current consumption measurement, a subsystem bandwidth usage measurement, and a subsystem efficiency measurement. The exemplary system may also include: means for determining an IPDC value using a predetermined relationship between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement. The exemplary system may also include: means for powering off the subsystem in response to a determination that the subsystem has been idle for an interval not less than a determined IPDC value.

[0011] An exemplary computer-readable medium for controlling powering off of a subsystem of a computing device may be provided. The computer-readable medium may include a non-transitory computer-readable medium having instructions stored thereon in a computer-executable form. These instructions, when executed by a processing system of a computing device, may configure the processing system to determine a subsystem current consumption measurement, a subsystem bandwidth usage measurement, and a subsystem efficiency measurement. These instructions may further configure the processing system to determine an IPDC value using a predetermined relationship between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement. These instructions may also configure the processing system to power off the subsystem in response to a determination of whether the subsystem is idle for an interval not less than a determined IPDC value. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] 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 "101A" or "101B", the letter character designation may distinguish between two similar parts or elements in the same figure. The letter character designation of a reference numeral may be omitted when it is intended that the reference numeral encompass all parts having the same reference numeral in all figures.

[0013] Figure 1 is a block diagram illustrating a system for controlling powering off of a subsystem of a computing device according to an exemplary embodiment.

[0014] Figure 2 is a flow chart illustrating a method for controlling powering off of a subsystem of a computing device according to an exemplary embodiment.

[0015] Figure 3 is a flow chart illustrating another method for controlling powering off of a subsystem of a computing device according to an exemplary embodiment.

[0016] Figure 4 is a graph of Idle Power Down Count ("IPDC") versus exemplary subsystem current consumption data points.

[0017] Figure 5 is a graph of exemplary subsystem current consumption versus exemplary subsystem bandwidth data points.

[0018] Figure 6 is a graph of exemplary subsystem bandwidth versus exemplary subsystem efficiency data points.

[0019] Figure 7 is a block diagram of a portable computing device according to an exemplary embodiment. DETAILED DESCRIPTION

[0020] The word "exemplary" is used herein to mean "serving as an example, instance, or illustration." The word "exemplary" may be used herein synonymously with "exemplary." Any aspect described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other aspects.

[0021] like Figure 1As shown, in an exemplary or exemplary embodiment, system 100 may include processor subsystem 102, memory controller 104, and memory 106. Memory 106 may be, for example, dynamic random access memory ("DRAM"), such as double data rate synchronous DRAM ("DDR-SDRAM" or simply "DDR"). Thus, memory controller 104 may be a DRAM (e.g., DDR) controller. Memory 106 and memory controller 104 may be collectively referred to as a memory subsystem. The memory subsystem may be a portable computing device or "PCD" ( Figure 1 The present invention may be one of any number of subsystems (not shown), which may be, for example, a laptop or palmtop computer, a cellular or smart phone, a personal digital assistant, a portable game console, a satellite phone, etc.

[0022] As described below, the system 100 may be configured to power down the memory subsystem based on the degree to which the memory subsystem is idle or inactive. The system 100 may include a power controller 108 that may be configured to provide a selected power supply voltage level to the memory controller 104 and the memory 106 via one or more power rails 110. As described below, powering down the memory subsystem may include the power controller 108 reducing the voltage level on the power rails 110. The system 100 may also be configured to power up the memory subsystem after it has been powered down based on the degree to which the memory subsystem is no longer idle or inactive. Powering up the memory subsystem may include the power controller 108 increasing the voltage level on the power rails 110.

[0023] Despite Figure 1 In the exemplary embodiment illustrated in , system 100 may be configured to power down a memory subsystem based on the idleness of the memory subsystem, but in other embodiments (not shown), such a system may be configured to power down a subsystem based on the idleness of a different subsystem. A system according to the principles described herein may be configured to power down any one or more subsystems that consume power and are sometimes idle. How idle is determined may depend on the type of subsystem, as understood by one of ordinary skill in the art. In some examples, whether a subsystem is "idle" may be based on whether an indicator of activity exceeds a threshold. In addition, unless otherwise indicated, the term "power down" may include a subsystem entering any type of low power or power saving mode, and is not limited to a subsystem being completely powered down as a whole.

[0024] The memory controller 104 may include an activity-based power manager (“APM”) 112. The APM 112 may include an idle power-down count (“IPDC”) register 114 and a counter 116. Since the memory controller 104 may include a processor or processing logic (not separately shown), the APM 112 may be implemented by executing software on such processing logic. However, in some examples, some or all of the APM 112 (such as the IPDC register 114, the counter 116, etc.) may be implemented in hardware. The counter 116 may be configured to be loaded or otherwise provided with an initial count value from the IPDC register 114, and the counter 116 may be configured to count down from the initial count value (e.g., to zero). When the counter 116 completes a full count (e.g., completes a countdown from the initial count value to zero), the counter 116 signals the power controller 108. In response to this signal indicating that the counter 116 has completed a full count, the power controller 108 may power down the memory subsystem. It should be understood that the counter 116 and IPDC register 114 together configured to count down from the count value stored in the IPDC register 114 are merely examples of how such counting logic may be implemented. For example, in other embodiments, such counting logic may count up to a provided count value rather than counting down from a provided count value.

[0025] The APM 112 may be configured to monitor the activity (or conversely, idleness) of the memory subsystem. If the APM 112 determines that the memory subsystem is active, i.e., not idle, the APM 112 may reload the counter 116 with the initial count value in the IPDC register 114. For example, the APM 112 may determine whether the subsystem is active by monitoring DRAM commands generated by the memory controller 104. It should be understood that in addition to being configured to perform or otherwise control the functions described herein with respect to powering down the memory subsystem, the memory controller 104 may be configured to perform or otherwise control conventional functions with respect to the storage of data in the memory 106, such as writing data to the memory 106 and reading data from the memory 106. The memory controller 104 may control writing data to the memory 106 and reading data from the memory 106 in response to memory transaction requests received from a processor (such as, for example, the processor subsystem 102). As understood by those of ordinary skill in the art, in response to such memory transaction requests, the memory controller 104 may generate DRAM commands and provide the DRAM commands to the DRAM 106.

[0026] In the illustrated embodiment, in the powered-down state, the memory subsystem may no longer be enabled to complete memory transactions. However, in some embodiments, there may be a low-power "power-down" state in which the memory subsystem may still be enabled to perform some operations. For example, in such a low-power "power-down" state, the memory subsystem may be enabled to perform memory transactions, but to a more limited extent than in a higher-power "power-up" state. Alternatively or in addition, in such a low-power "power-down" state, the memory subsystem may continue to perform refresh operations to maintain stored data but may not be enabled to perform memory transactions.

[0027] In an exemplary embodiment, the APM 112 may determine whether the memory subsystem is active by monitoring the DRAM commands generated by the memory controller 104. For example, the APM 112 may monitor a DRAM command commonly referred to as an activate or "ACT". When the APM 112 detects that the memory controller 104 has generated an ACT command, the APM 112 may reload the counter 116 with the initial count value in the IPDC register 114. Therefore, if the APM 112 detects the ACT command within an interval equal to the full count, the counter 116 will be reloaded with the initial count value and start counting down again from the initial count value. However, if the APM 112 does not detect the ACT command within an interval equal to the full count, the counter 116 will not be reloaded before reaching the full count, and therefore will signal the power controller 108 to power down the memory subsystem when the full count is reached. In other words, the memory controller 104 and the power controller 108 are configured together to power down the memory subsystem in response to determining that the memory subsystem has been idle for an interval of the full count.

[0028] If the initial count value is a constant or fixed value (which may be referred to as a "static IPDC"), it may be challenging to select a static IPDC that balances power savings and performance across a range of operating conditions. A higher (more conservative) static IPDC may result in higher performance at the expense of higher power consumption, while a lower (more aggressive) static IPDC may result in lower power consumption at the expense of lower performance. The solution described herein is related to determining the initial count value based on real-time operating conditions. That is, a "dynamic IPDC" may be determined based on the real-time operating conditions of the memory subsystem, rather than loading the counter 116 with a static IPDC as the initial count value. These conditions may include the amount of current consumed by the memory subsystem, the amount of bandwidth used by the memory subsystem, and the memory subsystem efficiency.

[0029] In the illustrated embodiment, the processor subsystem 102 may be configured to determine the dynamic IPDC. However, in other embodiments, another processor (not shown), the memory controller 104 or a portion thereof (e.g., APM 112), or other one or more components may be configured to determine the dynamic IPDC. In the illustrated embodiment, the processor subsystem 102 may include an IPDC manager 118, a memory subsystem ("SS") current monitor 120, and a memory subsystem performance monitor 122. The IPDC manager 118, the memory subsystem performance monitor 122, and the memory subsystem current monitor 120 may be implemented by executing software on the processor subsystem 102. An analog-to-digital conversion circuit that converts an analog current measurement into a digital value may be provided, but is not separately shown for clarity.

[0030] The memory subsystem performance manager 122 may be configured to monitor traffic on the memory bus 124 between the memory controller 104 and the memory 106, and may be configured to detect DRAM commands and other memory bus traffic. Based on such detected DRAM commands, the memory subsystem performance manager 122 may be configured to alert the IPDC manager 118 whenever the memory subsystem performance manager 122 detects that a memory transaction is being performed. The memory subsystem performance manager 122 may be configured to measure or determine the amount of bandwidth being used by the memory subsystem for memory transactions. Determining the amount of bandwidth being used by the memory subsystem may include determining the amount of data per unit time (e.g., bytes per second) transferred to or from the memory 106 in association with the memory transaction. Whenever a memory transaction is detected, the memory subsystem performance manager 122 may determine this bandwidth, i.e., collect memory subsystem bandwidth usage measurements. The memory subsystem performance manager 122 may therefore collect a certain number (N) of memory subsystem bandwidth usage measurements on N memory transactions.

[0031] The memory subsystem performance manager 122 may also be configured to determine the memory subsystem efficiency during each of the N memory transactions described above. As will be appreciated by one of ordinary skill in the art, the memory subsystem efficiency may be measured or determined by dividing the number of memory data cycles during a transaction by the number of memory busy cycles during the transaction. The memory subsystem performance manager 122 may therefore collect N memory subsystem efficiency measurements over the N memory transactions.

[0032] The memory subsystem current monitor 120 may be configured to monitor the current on the power rail 110 that powers the memory controller 104 and the memory 106. The memory subsystem current monitor 120 may be configured to measure or determine the amount of current consumed by the memory controller 104 and the memory 106 during each of the N memory transactions described above. The memory subsystem current monitor 120 may therefore collect N current consumption measurements over the N memory transactions.

[0033] The IPDC manager 118 may determine the IPDC value in response to one or more of the memory current consumption measurements, one or more of the memory bandwidth usage measurements, and one or more of the memory subsystem efficiency measurements. In determining the IPDC value, the IPDC manager 118 may use a predetermined relationship between the memory subsystem current consumption, the subsystem bandwidth usage, and the subsystem efficiency. The predetermined relationship may include, for example, a first relationship between IPDC and the memory subsystem current consumption, a second relationship between the memory subsystem current consumption measurement and the memory subsystem bandwidth usage, and a third relationship between the memory subsystem bandwidth usage and the memory subsystem efficiency. The first relationship, the second relationship, and the third relationship may be, for example, polynomial equations.

[0034] The first relationship may be, for example, the following fifth-order polynomial equation:

[0035] y=a 1 x 5 +a 2 x 4 +a 3 x 3 +a 4 x 2 +a 5 x+c 1 , (Equation 1)

[0036] Where x is the IPDC value, y is the memory subsystem current consumption, and a 1 、a 2 、a 3 、a 4 and a 5 is the coefficient, and c 1 is a constant.

[0037] The second relationship may be, for example, the following fourth-order polynomial equation:

[0038] z=b 1 y 4 +b 2 y 3 +b 3 y 2 +b 4 y+c 2 , (Equation 2)

[0039] where z is the memory subsystem bandwidth usage, b 1 、b 2 、b 3 and b 4 is the coefficient, and c 2 is a constant.

[0040] The third relationship may be, for example, the following third-order polynomial equation:

[0041] p=d 1 z 3 +d 2 z 2 +d 3 z+c 3 , (Equation 3)

[0042] where p is the memory subsystem efficiency, d 1 d 2 and d 3 is the coefficient, and c 3 is a constant.

[0043] Determining the IPDC value x using the three polynomial equations described above may include determining a memory subsystem current consumption value (i.e., a value of y), a memory subsystem bandwidth usage value (i.e., a value of z), and a memory subsystem efficiency value (i.e., a value of p), and solving a system of three polynomial equations for x using those values ​​of y, z, and p. As will be appreciated by one of ordinary skill in the art, any of a variety of known techniques may be used to solve such a system of polynomial equations. The Newton-Raphson method is an example of one such technique.

[0044] The IPDC manager 118 may be configured to determine an IPDC value every N memory transactions (i.e., at intervals of N memory transactions). Determining the IPDC value may include: using N memory subsystem bandwidth usage measurements, N memory subsystem efficiency measurements, and N memory subsystem current consumption measurements to determine the IPDC value. For example, the IPDC manager 118 may use a histogram method to determine the most common (i.e., most frequently occurring) memory subsystem bandwidth usage measurement among the N memory subsystem bandwidth usage measurements, the most common memory subsystem efficiency measurement among the N memory subsystem efficiency measurements, and the most common memory subsystem current consumption measurement among the N memory subsystem current consumption measurements. Then, using the most common memory subsystem current consumption measurement as the above-mentioned memory subsystem current consumption value y, using the most common memory subsystem bandwidth measurement as the above-mentioned memory subsystem bandwidth value z, and using the most common memory subsystem efficiency measurement as the memory subsystem efficiency value p, the IPDC manager 118 may use a predetermined relationship between the memory subsystem current consumption, the memory subsystem bandwidth usage, and the memory subsystem efficiency to determine the IPDC value x. For example, IPDC manager 118 may be configured to solve the system of three polynomial equations described above to obtain IPDC value x. IPDC manager 118 may then program IPDC register 114 with IPDC value x.

[0045] Although in the exemplary embodiment ( Figure 1 ), the IPDC manager 118 of the processor subsystem 102 is configured to solve the system of three polynomial equations described above to determine the IPDC value x, but in other embodiments (not shown), another processing subsystem may determine the IPDC value, or the IPDC value may be determined in another manner. For example, the APM 112 or other parts of the memory controller 104 may be configured to calculate or otherwise determine the IPDC value.

[0046] exist Figure 2 , a method 200 for controlling power off of a subsystem of a portable computing device or "PCD" is illustrated. The method 200 may include the following. As indicated by box 202, the method 200 may include: determining a subsystem current consumption measurement, a subsystem bandwidth usage measurement, and a subsystem efficiency measurement. The above-mentioned IPDC manager 118, memory subsystem current monitor 120, and memory subsystem performance monitor 122 may be configured to control these aspects of the method 200. The IPDC manager 118 and the memory subsystem current monitor 120 may be examples of components for determining a subsystem current consumption measurement. The IPDC manager 118 and the memory subsystem performance monitor 122 may be examples of components for determining a subsystem bandwidth usage measurement and a subsystem efficiency measurement.

[0047] As indicated at block 204, method 200 may include determining an IPDC value using a predetermined relationship between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to one or more subsystem current consumption measurements, one or more subsystem bandwidth usage measurements, and one or more subsystem efficiency measurements. IPDC manager 118 may be an example of a component for determining an IPDC value using a predetermined relationship between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to subsystem current consumption measurements, subsystem bandwidth usage measurements, and subsystem efficiency measurements.

[0048] As indicated at block 206, method 200 may include powering down a subsystem based on determining whether the subsystem is idle for an interval not less than the IPDC value determined (block 204). APM 112 and power controller 108 may be examples of means for powering down a subsystem in response to determining whether the subsystem is idle for an interval not less than the determined IPDC value. Although in the illustrated embodiment ( Figure 1 ), the processor subsystem 102 and the memory controller 104 are configured to execute or control the method 200, but in other embodiments, other types of processing systems may be configured to control the method 200.

[0049] exist Figure 3 , a method 300 for controlling power-off of a subsystem of a PCD is illustrated. The method 300 may include the following. As indicated by block 302, the method 300 may begin by initializing the IPDC register 114 (“static_IPDC”) by storing a static IPDC value (“static_IPDC”) in the IPDC register 114. Figure 1 ). The static IPDC value is a constant or fixed value that does not change during operation of the PCD. As indicated at block 304, method 300 may include obtaining, for each of N memory (e.g., DRAM) transactions: a DRAM subsystem current consumption measurement, a DRAM subsystem bandwidth usage measurement, and a DRAM subsystem efficiency measurement.

[0050] As indicated at block 306, method 300 may include determining a most common (i.e., most frequently occurring) DRAM subsystem current consumption measurement among the N DRAM subsystem current consumption measurements, a most common DRAM subsystem bandwidth usage measurement among the N DRAM subsystem bandwidth usage measurements, and a most common DRAM subsystem efficiency measurement among the N DRAM subsystem efficiency measurements. Determining these most common measurements or values ​​may include using a histogram technique.

[0051] As indicated at block 308, method 300 may include calculating a dynamic IPDC value ("dynamic_IPDC") based on the most common values ​​of DRAM subsystem current consumption, DRAM subsystem bandwidth usage, and DRAM subsystem efficiency determined from block 306. Calculating the dynamic IPDC value may include solving a system of polynomial equations, as described above.

[0052] As indicated by block 310, method 300 may include determining a change or increment between a calculated or dynamic IPDC value and a previous IPDC value: Δ=|dynamic_IPDC-previous_IPDC|. Note that the increment may be the absolute value of the difference between the dynamic IPDC value and the previous IPDC value. Initially (i.e., when the dynamic IPDC value is calculated for the first time according to block 308), the previous IPDC value may be the static IPDC value described above. (Initializing the value of previous_IPDC to the value of static_IPDC may be performed in association with block 302). For each subsequent calculation of a dynamic IPDC value, the calculated dynamic IPDC value may be assigned to the previous IPDC value, as indicated by block 312.

[0053] As indicated at block 314, method 300 may include determining whether the above-described delta between the calculated or dynamic IPDC value and the previous IPDC value is greater than a threshold value K. Threshold value K may be a constant or fixed value. If it is determined (block 314) that the delta is not greater than threshold value K, method 300 may return to block 304 described above, where measurements may again be obtained over N DRAM transactions. The actions described above with respect to blocks 304 to 314 may be repeated for every N DRAM transactions.

[0054] If it is determined (block 314) that the increment is greater than the threshold value K, the IPDC register 114 ( ) may be updated by storing a calculated or dynamic IPDC value ("dynamic_IPDC") in the IPDC register 114. Figure 1 ) is reprogrammed, as indicated by block 316. The method 300 may then return to block 304 described above.

[0055] exist Figure 4 , an exemplary graph 400 of IPDC value versus DDR subsystem current consumption is illustrated. Graph 400 shows (in solid line) a curve 402 fitted to data points 404. Graph 400 also shows (in dashed line) a curve 406, which is a polynomial approximation of curve 402. In the illustrated example, the polynomial equation may be:

[0056] y=1E -8 x 5 -9E -6 x 4 +0.0025x 3-0.2494x 2 -8.1243x+597.7 (Equation 4)

[0057] In equation 4, the coefficient 1E -8 is an example of coefficient a1 in equation 1 above, coefficient -9E -6 is an example of coefficient a2 in equation 1, coefficient 0.0025 is an example of coefficient a3 in equation 1, coefficient -0.2494 is an example of coefficient a4 in equation 1, coefficient -8.1243 is an example of coefficient a5 in equation 1, and number 597.7 is an example of constant c1 in equation 1. Data points 404 may be collected prior to operating the PDC, and the polynomial approximation may be determined from data points 404. For example, a manufacturer of a PCD or a portion thereof (e.g., a system on a chip including IPDC manager 118) may collect data points 404 and determine the polynomial approximation prior to providing the PCD to an end user. The DDR subsystem may be operated in data collection mode independent of other PCD functions (e.g., using a software tool controlled by a human operator). Data points 404 may be collected by operating the DDR subsystem over multiple transactions. The data collection mode may enable direct access (via a software tool) to modify IPDC registers 114 ( Figure 1 ) to bypass the IPDC manager 118. Figure 1 ) may vary with those transactions, and the corresponding DDR subsystem current consumption may be measured in each of those transactions to provide data point 504.

[0058] exist Figure 5 , an exemplary graph 500 of DDR subsystem current consumption versus DDR subsystem bandwidth usage is illustrated. Graph 500 shows (in a solid line) a curve 502 that is fitted to data points 504. Graph 500 also shows (in a dashed line) a curve 506 that is a polynomial approximation of curve 502. In the illustrated example, the polynomial equation may be:

[0059] z=9E -8 y 4 -0.0003y 3 +0.286y 2 -136.28y+38983 (Equation 5)

[0060] In equation 5, the coefficient 9E -8is an example of coefficient b1 in Equation 2 above, coefficient -0.0003 is an example of coefficient b2 in Equation 2, coefficient 0.286 is an example of coefficient b3 in Equation 2, coefficient -136.28 is an example of coefficient b4 in Equation 2, and the number 38983 is an example of constant c2 in Equation 2. Data points 504 may be collected prior to operating the PDC, and the polynomial approximation may be determined from data points 504. Data points 504 may be collected by operating the DDR subsystem over multiple transactions, as described above. Programmed into IPDC register 114 ( Figure 1 ) may vary with those transactions, and the DDR subsystem current consumption and DDR subsystem bandwidth, which together provide data point 504, may be measured in each of those transactions.

[0061] exist Figure 6 , an exemplary graph 600 of DDR subsystem bandwidth usage versus DDR subsystem efficiency is illustrated. Graph 600 shows (in a solid line) a curve 602 fitted to data points 604. Graph 600 also shows (in a dashed line) a curve 606, which is a polynomial approximation of curve 602. In the illustrated example, the polynomial equation may be:

[0062] z=7E -5 y 3 -3.0718y 2 +45619y-2E 8 (Equation 6)

[0063] In equation 6, the coefficient 7E -5 is an example of coefficient d1 in the above equation 3, coefficient -3.0718 is an example of coefficient d2 in equation 3, coefficient 45619 is an example of coefficient d3 in equation 3, and the number -2E 8 is an example of constant c3 in Equation 3. Data points 604 may be collected prior to operating the PDC and the polynomial approximation may be determined from data points 604. Data points 604 may be collected by operating the DDR subsystem over multiple transactions, as described above. Figure 1 ) may vary with those transactions, and the DDR subsystem bandwidth and DDR subsystem efficiency provided along with data point 604 may be measured in each of those transactions.

[0064] After the polynomial approximation (ie, equation) is determined in the manner described above, its coefficients and constants may be stored for use by the IPDC manager 118 ( Figure 1) for later use. The polynomial approximation (or its coefficients and constants) may be stored, for example, in a non-volatile memory (not shown) of the processor subsystem 102 that is accessible by the IPDC manager 118. Subsequently, when the PCD is being operated by an end user in its normal operating mode, the IPDC manager 118 may access the stored polynomial equation information to perform the method 200 (described above). Figure 2 ) and 300( Figure 3 ) as described above.

[0065] Figure 7 An example of a PCD 700 such as a mobile phone or smart phone is illustrated in which exemplary embodiments of systems, methods, computer-readable media, and other examples of controlling powering down of a PDC subsystem may be provided. Other examples of PCDs (not shown) in which embodiments of such systems, methods, computer-readable media, etc. may be provided include laptop computers, tablet computers, smartbooks, satellite phones, navigation devices, portable game consoles, personal digital assistants ("PDAs"), etc. For purposes of clarity, some data buses, interconnects, signals, etc. are not shown in the figure. Figure 7 Shown in.

[0066] PCD 700 may include SoC 702. SoC 702 may include a central processing unit ("CPU") 704, a neural processing unit ("NPU") 705, a graphics processing unit ("GPU") 706, a digital signal processor ("DSP") 707, an analog signal processor 708, a modem / modem subsystem 754, or other processors. CPU 704 may include one or more CPU cores, such as a first CPU core 704A, a second CPU core 704B, and so on to an Nth CPU core 704N. CPU 704 may be the CPU core described above. Figure 1 An example of processor subsystem 102 is shown.

[0067] The PCD 700 may include a display controller 710 and a touch screen controller 712 coupled to the CPU 704. A touch screen display 714 external to the SoC 702 may be coupled to the display controller 710 and the touch screen controller 712. The PCD 700 may also include a video decoder 716 coupled to the CPU 704. A video amplifier 718 may be coupled to the video decoder 716 and the touch screen display 714. A video port 720 may be coupled to the video amplifier 718. A universal serial bus (“USB”) controller 722 may also be coupled to the CPU 704, and a USB port 724 may be coupled to the USB controller 722. A subscriber identity module (“SIM”) card 726 may also be coupled to the CPU 704.

[0068] One or more memories may be coupled to CPU 704. The one or more memories may include both volatile memory and non-volatile memory. Examples of volatile memory include static random access memory ("SRAM") 728 and dynamic random access memory ("DRAM") 730 and 731. Such memories may be located external to SoC 702 (such as DRAM 730) or internal to SoC 702 (such as DRAM 731). A DRAM controller 732 coupled to CPU 704 may control writing data to and reading data from DRAM 730 and 731. DRAM controller 732 may be a memory controller described above with respect to DRAM 730 and 731. Figure 1 The memory controller 104 is described as an example, and the DRAM 730 or 731 may be an example of the memory 106.

[0069] A stereo audio codec 734 may be coupled to the analog signal processor 708. Further, an audio amplifier 736 may be coupled to the stereo audio codec 734. A first stereo speaker 738 and a second stereo speaker 740 may be coupled to the audio amplifier 736, respectively. In addition, a microphone amplifier 742 may be coupled to the stereo audio codec 734, and a microphone 744 may be coupled to the microphone amplifier 742. A frequency modulation ("FM") radio tuner 746 may be coupled to the stereo audio codec 734. An FM antenna 748 may be coupled to the FM radio tuner 746. Further, a stereo headset 750 may be coupled to the stereo audio codec 734. Other devices that may be coupled to the CPU 704 include one or more digital (e.g., CCD or CMOS) cameras 752.

[0070] A modem or RF transceiver 754 may be coupled to the analog signal processor 708 and the CPU 704. An RF switch 756 may be coupled to the RF transceiver 754 and the RF antenna 758. In addition, a keypad 760, a mono headset 762 with a microphone, and a vibrator device 764 may be coupled to the analog signal processor 708.

[0071] SoC 702 may have one or more internal or on-chip thermal sensors 770A and may be coupled to one or more external or off-chip thermal sensors 770B. Analog-to-digital converter controller 772 may convert the voltage drop generated by thermal sensors 770A and 770B into a digital signal. Power supply 774 and PMIC 776 may provide power to SoC 702.

[0072] The firmware or software may be stored in any of the above memories, such as DRAM 730 or 731, SRAM 728, etc., or may be stored in local memory directly accessible by 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.

[0073] Specific implementation examples are described in the following numbered clauses.

[0074] 1. A method for controlling power-off of a subsystem of a computing device, comprising:

[0075] determining, by the processing system, a subsystem current consumption measure, a subsystem bandwidth usage measure, and a subsystem efficiency measure;

[0076] determining, by the processing system in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement, an idle power-down count value using a plurality of predetermined relationships between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency; and

[0077] The subsystem is powered down in response to a determination of whether the subsystem has been idle for an interval that is not less than a determined idle power down count value.

[0078] 2. The method of clause 1, wherein determining the idle power-down count value comprises: determining a most common subsystem current consumption measurement from a plurality of subsystem current consumption measurements, determining a most common subsystem bandwidth usage measurement from a plurality of subsystem bandwidth usage measurements,

[0079] and identifying the most common subsystem efficiency measure from among multiple subsystem efficiency measures.

[0080] 3. The method of clause 1 or 2, wherein the subsystem comprises a dynamic random access memory subsystem.

[0081] 4. A method according to any one of clauses 1 to 3, wherein the plurality of predetermined relationships include a first relationship between an idle power-down count and subsystem current consumption, a second relationship between a subsystem current consumption measurement and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency.

[0082] 5. A method according to any one of clauses 1 to 4, wherein the first relationship is a fifth-order polynomial equation, the second relationship is a fourth-order polynomial equation, and the third relationship is a third-order polynomial equation, and determining the idle power-down count value includes solving a system of polynomial equations.

[0083] 6. The method of any one of clauses 1 to 5, wherein powering down the subsystem comprises:

[0084] determining whether a difference between the determined idle power-off count value and a previous idle power-off count value exceeds a threshold;

[0085] When the subsystem is idle for an interval not less than the determined idle power-down count value and when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold, powering off the subsystem; and

[0086] The subsystem is powered off when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold value.

[0087] 7. The method of clause 6, wherein powering off the subsystem further comprises:

[0088] providing the determined idle power-down count value to a counting logic component when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold;

[0089] When the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold, providing the previous idle power-down count value to the counting logic component;

[0090] When the subsystem is active, resetting the counting logic; and

[0091] The subsystem is powered down when the counting logic counts a set number of clock cycles while the subsystem is inactive.

[0092] 8. A system for controlling power off of a subsystem of a computing device, comprising:

[0093] A processing system, the processing system being configured to:

[0094] Determine subsystem current consumption measurements, subsystem bandwidth usage measurements, and subsystem

[0095] System efficiency measurement;

[0096] In response to the subsystem current consumption measurement, the subsystem bandwidth usage

[0097] determining an idle power-down count value using a plurality of predetermined relationships between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency by measuring and measuring said subsystem efficiency; and

[0098] A power controller is configured to power off the subsystem in response to a determination whether the subsystem has been idle for an interval that is not less than a determined idle power-off count value.

[0099] 9. A system according to claim 8, wherein the processing system is configured to determine the determined idle power-off count value by being configured to perform the following operations: determining a most common subsystem current consumption measurement from multiple subsystem current consumption measurements, determining a most common subsystem bandwidth usage measurement from multiple subsystem bandwidth usage measurements, and determining a most common subsystem efficiency measurement from multiple subsystem efficiency measurements.

[0100] 10. The system of clause 8 or 9, wherein the subsystem comprises a dynamic random access memory subsystem.

[0101] 11. A system according to any one of clauses 8 to 10, wherein the plurality of predetermined relationships include a first relationship between idle power-down count and subsystem current consumption, a second relationship between subsystem current consumption measurement and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency.

[0102] 12. A system according to any one of clauses 8 to 11, wherein the first relationship is a fifth-order polynomial equation, the second relationship is a fourth-order polynomial equation, and the third relationship is a third-order polynomial equation, and determining the determined idle power-down count value includes solving a system of polynomial equations.

[0103] 13. The system of any of clauses 8 to 12, wherein the processing system and the power controller are configured to power down the subsystem by being configured to:

[0104] determining whether a difference between the determined idle power-off count value and a previous idle power-off count value exceeds a threshold;

[0105] When the subsystem is idle for an interval not less than the determined idle power-down count value and when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold, powering off the subsystem; and

[0106] The subsystem is powered off when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold value.

[0107] 14. The system of clause 13, wherein the processing system and the power controller are configured to power down the subsystem by being further configured to:

[0108] providing the determined idle power-down count value to a counting logic component when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold;

[0109] When the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold, providing the previous idle power-down count value to the counting logic component;

[0110] When the subsystem is active, resetting the counting logic; and

[0111] The subsystem is powered down when the counting logic counts a set number of clock cycles while the subsystem is inactive.

[0112] 15. The system of any of clauses 8 to 14, wherein the computing device comprises one of: a smartphone, a laptop, a tablet, a smartbook, a satellite phone, a navigation device, a portable game console, or a personal digital assistant.

[0113] 16. A system for controlling power off of a subsystem of a computing device, comprising:

[0114] means for determining a subsystem current consumption measure, a subsystem bandwidth usage measure, and a subsystem efficiency measure;

[0115] means for determining an idle power-down count value using a plurality of predetermined relationships among subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement; and

[0116] Means for powering off the subsystem in response to a determination whether the subsystem has been idle for an interval not less than a determined idle power-off count value.

[0117] 17. A system according to clause 16, wherein the component for determining the idle power-down count value includes a component for performing the following operations: determining a most common subsystem current consumption measurement from a plurality of subsystem current consumption measurements, determining a most common subsystem bandwidth usage measurement from a plurality of subsystem bandwidth usage measurements, and determining a most common subsystem efficiency measurement from a plurality of subsystem efficiency measurements.

[0118] 18. The system of clause 16 or 17, wherein the subsystem comprises a dynamic random access memory subsystem.

[0119] 19. A system according to any one of clauses 16 to 18, wherein the plurality of predetermined relationships include a first relationship between idle power-down counts and subsystem current consumption, a second relationship between subsystem current consumption measurements and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency.

[0120] 20. A system according to any of clauses 16 to 19, wherein the first relationship is a fifth-order polynomial equation, the second relationship is a fourth-order polynomial equation, and the third relationship is a third-order polynomial equation, and determining the idle power-down count value includes solving a system of polynomial equations.

[0121] 21. The system of any of clauses 16 to 20, wherein the means for powering down the subsystem comprises:

[0122] means for determining whether a difference between the determined idle power-down count value and a previous idle power-down count value exceeds a threshold;

[0123] A component for performing the following operations: when the subsystem is idle for an interval not less than the determined idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value exceeds the threshold, powering off the subsystem; and when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold, powering off the subsystem.

[0124] 22. The system of clause 21, wherein the means for powering down the subsystem further comprises:

[0125] means for providing said determined idle power-down count value to a counting logic component when said subsystem difference between said determined idle power-down count value and said previous idle power-down count value exceeds said threshold;

[0126] means for providing the previous idle power-down count value to the counting logic component when the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold value;

[0127] means for resetting said counting logic when said subsystem is active; and

[0128] Means for powering down the subsystem when the counting logic counts a set number of clock cycles while the subsystem is inactive.

[0129] 23. The system of any of clauses 16 to 22, wherein the computing device comprises one of: a smartphone, a laptop, a tablet, a smartbook, a satellite phone, a navigation device, a portable game console, or a personal digital assistant.

[0130] 24. A computer-readable medium for controlling powering off a subsystem of a computing device, the computer-readable medium comprising a non-transitory computer-readable medium having instructions stored thereon in a computer-executable form, the instructions, when executed by a processing system of the computing device, configuring the processing system to:

[0131] determining a subsystem current consumption measure, a subsystem bandwidth usage measure, and a subsystem efficiency measure;

[0132] determining an idle power-down count value using a plurality of predetermined relationships between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement; and

[0133] The subsystem is powered down in response to a determination of whether the subsystem has been idle for an interval that is not less than a determined idle power down count value.

[0134] 25. The computer-readable medium of clause 24, wherein the instructions for configuring the processing system to determine the idle power-down count value configure the processing system to: determine a most common subsystem current consumption measurement from a plurality of subsystem current consumption measurements, determine a most common subsystem bandwidth usage measurement from a plurality of subsystem bandwidth usage measurements,

[0135] and identifying the most common subsystem efficiency measure from among multiple subsystem efficiency measures.

[0136] 26. The computer-readable medium of clause 24 or 25, wherein the subsystem comprises a dynamic random access memory subsystem.

[0137] 27. A computer-readable medium according to any one of clauses 24 to 26, wherein the plurality of predetermined relationships include a first relationship between an idle power-down count and subsystem current consumption, a second relationship between a subsystem current consumption measurement and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency.

[0138] 28. A computer-readable medium according to any one of clauses 24 to 27, wherein the first relationship is a fifth-order polynomial equation, the second relationship is a fourth-order polynomial equation, and

[0139] The third relationship is a third order polynomial equation, and determining the idle power down count value includes solving a system of polynomial equations.

[0140] 29. The computer-readable medium of any of clauses 24 to 28, wherein the instructions to configure the processing system to power down the subsystem include instructions to configure the processing system to:

[0141] determining whether a difference between the determined idle power-off count value and a previous idle power-off count value exceeds a threshold;

[0142] When the subsystem is idle for an interval not less than the determined idle power-down count value and when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold, powering off the subsystem; and

[0143] The subsystem is powered off when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold value.

[0144] 30. The computer-readable medium of clause 29, wherein the instructions to configure the processing system to power down the subsystem further comprise instructions to configure the processing system to:

[0145] providing the determined idle power-down count value to a counting logic component when the subsystem difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold;

[0146] When the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold, providing the previous idle power-down count value to the counting logic component;

[0147] When the subsystem is active, resetting the counting logic; and

[0148] The subsystem is powered down when the counting logic counts a set number of clock cycles while the subsystem is inactive.

[0149] Alternative embodiments will become apparent to those of ordinary skill in the art.Thus, while selected aspects have been illustrated and described in detail, it should be understood that various substitutions and changes may be made therein.

Claims

1. A method for controlling power-off of a subsystem of a computing device, comprising: determining, by the processing system, a subsystem current consumption measure, a subsystem bandwidth usage measure, and a subsystem efficiency measure; determining, by the processing system in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement, an idle power-down count value using a plurality of predetermined relationships between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency; as well as The subsystem is powered down in response to a determination of whether the subsystem has been idle for an interval that is not less than a determined idle power down count value.

2. The method according to claim 1, wherein determining the idle power-down count value comprises: A most common subsystem current consumption measurement is determined from the plurality of subsystem current consumption measurements, a most common subsystem bandwidth usage measurement is determined from the plurality of subsystem bandwidth usage measurements, and a most common subsystem efficiency measurement is determined from the plurality of subsystem efficiency measurements.

3. The method of claim 1, wherein the subsystem comprises a dynamic random access memory subsystem.

4. The method of claim 3, wherein the plurality of predetermined relationships include a first relationship between idle power-down counts and subsystem current consumption, a second relationship between subsystem current consumption measurements and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency. 5 . The method of claim 4 , wherein the first relationship is a fifth-order polynomial equation, the second relationship is a fourth-order polynomial equation, and the third relationship is a third-order polynomial equation, and determining the idle power-down count value comprises solving a system of polynomial equations.

6. The method of claim 3, wherein powering off the subsystem comprises: determining whether a difference between the determined idle power-off count value and a previous idle power-off count value exceeds a threshold; When the subsystem is idle for an interval not less than the determined idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value exceeds the threshold, powering off the subsystem; as well as The subsystem is powered off when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold value.

7. The method of claim 6, wherein powering off the subsystem further comprises: providing the determined idle power-down count value to a counting logic component when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold; When the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold, providing the previous idle power-down count value to the counting logic component; When the subsystem is active, resetting the counting logic component; as well as The subsystem is powered down when the counting logic counts a set number of clock cycles while the subsystem is inactive.

8. A system for controlling power off of a subsystem of a computing device, comprising: A processing system, the processing system being configured to: determining a subsystem current consumption measure, a subsystem bandwidth usage measure, and a subsystem efficiency measure; determining an idle power-down count value using a plurality of predetermined relationships between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement; and A power controller is configured to power off the subsystem in response to a determination whether the subsystem has been idle for an interval that is not less than a determined idle power-off count value.

9. The system of claim 8, wherein the processing system is configured to determine the determined idle power-down count value by being configured to: determine a most common subsystem current consumption measurement from a plurality of subsystem current consumption measurements, determine a most common subsystem bandwidth usage measurement from a plurality of subsystem bandwidth usage measurements, and determine a most common subsystem efficiency measurement from a plurality of subsystem efficiency measurements.

10. The system of claim 8, wherein the subsystem comprises a dynamic random access memory subsystem.

11. The system of claim 10, wherein the plurality of predetermined relationships include a first relationship between idle power-down counts and subsystem current consumption, a second relationship between subsystem current consumption measurements and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency.

12. The system of claim 11, wherein the first relationship is a fifth order polynomial equation, the second relationship is a fourth order polynomial equation, and the third relationship is a third order polynomial equation, and determining the determined idle power-down count value comprises solving a system of polynomial equations.

13. The system of claim 10, wherein the processing system and the power controller are configured to power down the subsystem by being configured to: determining whether a difference between the determined idle power-off count value and a previous idle power-off count value exceeds a threshold; When the subsystem is idle for an interval not less than the determined idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value exceeds the threshold, powering off the subsystem; as well as The subsystem is powered off when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold value.

14. The system of claim 13, wherein the processing system and the power controller are configured to power down the subsystem by being further configured to: providing the determined idle power-down count value to a counting logic component when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold; When the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold, providing the previous idle power-down count value to the counting logic component; When the subsystem is active, resetting the counting logic component; as well as The subsystem is powered down when the counting logic counts a set number of clock cycles while the subsystem is inactive.

15. The system of claim 8, wherein the computing device comprises one of: a smartphone, a laptop, a tablet, a smartbook, a satellite phone, a navigation device, a portable game console, or a personal digital assistant.

16. A system for controlling power off of a subsystem of a computing device, comprising: means for determining a subsystem current consumption measure, a subsystem bandwidth usage measure, and a subsystem efficiency measure; means for determining an idle power-down count value using a plurality of predetermined relationships between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement; and Means for powering off the subsystem in response to a determination whether the subsystem has been idle for an interval not less than a determined idle power-off count value.

17. The system of claim 16, wherein the means for determining the idle power-down count value comprises means for determining a most common subsystem current consumption measurement from a plurality of subsystem current consumption measurements, determining a most common subsystem bandwidth usage measurement from a plurality of subsystem bandwidth usage measurements, and determining a most common subsystem efficiency measurement from a plurality of subsystem efficiency measurements.

18. The system of claim 16, wherein the subsystem comprises a dynamic random access memory subsystem.

19. The system of claim 18, wherein the plurality of predetermined relationships include a first relationship between idle power-down counts and subsystem current consumption, a second relationship between subsystem current consumption measurements and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency.

20. The system of claim 19, wherein the first relationship is a fifth order polynomial equation, the second relationship is a fourth order polynomial equation, and the third relationship is a third order polynomial equation, and determining the idle power-down count value comprises solving a system of polynomial equations.

21. The system of claim 18, wherein the means for powering down the subsystem comprises: means for determining whether a difference between the determined idle power-down count value and a previous idle power-down count value exceeds a threshold; means for: powering off the subsystem when the subsystem is idle for an interval not less than the determined idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value exceeds the threshold; and powering off the subsystem when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold.

22. The system of claim 21, wherein the means for powering off the subsystem further comprises: means for providing said determined idle power-down count value to a counting logic component when said difference between said determined idle power-down count value and said previous idle power-down count value exceeds said threshold value; means for providing the previous idle power-down count value to the counting logic component when the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold value; means for resetting said counting logic when said subsystem is active; and Means for powering down the subsystem when the counting logic counts a set number of clock cycles while the subsystem is inactive.

23. The system of claim 16, wherein the computing device comprises one of: a smartphone, a laptop computer, a tablet computer, a smartbook, a satellite phone, a navigation device, a portable game console, or a personal digital assistant.

24. A computer-readable medium for controlling powering off a subsystem of a computing device, the computer-readable medium comprising a non-transitory computer-readable medium having instructions stored thereon in a computer-executable form, the instructions, when executed by a processing system of the computing device, configuring the processing system to: determining a subsystem current consumption measure, a subsystem bandwidth usage measure, and a subsystem efficiency measure; determining an idle power-down count value using a plurality of predetermined relationships between subsystem current consumption, subsystem bandwidth usage, and subsystem efficiency in response to the subsystem current consumption measurement, the subsystem bandwidth usage measurement, and the subsystem efficiency measurement; and The subsystem is powered down in response to a determination of whether the subsystem has been idle for an interval that is not less than a determined idle power down count value.

25. The computer-readable medium of claim 24, wherein the instructions for configuring the processing system to determine the idle power-down count value configure the processing system to: determine a most common subsystem current consumption measurement from a plurality of subsystem current consumption measurements, determine a most common subsystem bandwidth usage measurement from a plurality of subsystem bandwidth usage measurements, and determine a most common subsystem efficiency measurement from a plurality of subsystem efficiency measurements.

26. The computer-readable medium of claim 24, wherein the subsystem comprises a dynamic random access memory subsystem.

27. The computer-readable medium of claim 26, wherein the plurality of predetermined relationships comprises a first relationship between idle power-down counts and subsystem current consumption, a second relationship between subsystem current consumption measurements and subsystem bandwidth usage, and a third relationship between subsystem bandwidth usage and subsystem efficiency.

28. The computer-readable medium of claim 27, wherein the first relationship is a fifth-order polynomial equation, the second relationship is a fourth-order polynomial equation, and the third relationship is a third-order polynomial equation, and determining the idle power-down count value comprises solving a system of polynomial equations.

29. The computer-readable medium of claim 26, wherein the instructions to configure the processing system to power down the subsystem include instructions to configure the processing system to: determining whether a difference between the determined idle power-off count value and a previous idle power-off count value exceeds a threshold; When the subsystem is idle for an interval not less than the determined idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value exceeds the threshold, powering off the subsystem; as well as The subsystem is powered off when the subsystem is idle for an interval not less than the previous idle power-off count value and when the difference between the determined idle power-off count value and the previous idle power-off count value does not exceed the threshold value.

30. The computer-readable medium of claim 29, wherein the instructions to configure the processing system to power down the subsystem further comprise instructions to configure the processing system to: providing the determined idle power-down count value to a counting logic component when the difference between the determined idle power-down count value and the previous idle power-down count value exceeds the threshold; When the difference between the determined idle power-down count value and the previous idle power-down count value does not exceed the threshold, providing the previous idle power-down count value to the counting logic component; When the subsystem is active, resetting the counting logic component; as well as The subsystem is powered down when the counting logic counts a set number of clock cycles while the subsystem is inactive.

Citation Information

Patent Citations

  • Adaptive power-down of disk drives based on predicted idle time

    US20150100810A1

  • Dynamic power management optimization

    US20160378168A1

  • Systems and methods for dynamically adjusting memory state transition timers

    US20170068308A1