JPWO2013145294A1 - Information processing apparatus, power supply control method, and power supply control program - Google Patents

Abstract

It is an object to accurately estimate the timing when the power supplied from the power supply unit reaches the maximum power that can be supplied by the power supply unit. Power can be supplied to the system board (12) by the plurality of power supply units (14), and the current sensor (28) can change the system board (14) from the operating power supply unit (14) among the plurality of power supply units (14). A physical quantity corresponding to the power supplied to 12) is measured every predetermined time. Then, based on the history of measurement results obtained by the current sensor (28), the power supplied from the operating power supply unit (14) to the system board (12) is converted by the estimation unit (32) into the operating power supply unit (14). ) Estimate when to reach the maximum power supply.

Description

  The present invention relates to an information processing apparatus, a power supply control method, and a power supply control program.

  The server device includes a system board on which a CPU (central processing unit) and the like are mounted. As the scale of the server device increases, the number of system boards increases, and the number of CPUs increases accordingly. Therefore, power consumption also increases. Therefore, in order to supply power as much as possible to the operating CPU and the like as much as possible, a plurality of power supply units are provided, the CPU power consumption is predicted, and when the CPU power consumption tends to increase, a prediction is made. Measures may be taken to add one power supply unit to be operated according to power consumption.

  However, even if the power supply unit to be operated is additionally activated, it is necessary to consider the startup time of the power supply unit. In other words, since it takes a certain amount of time for the supplied power to become stable and operating after the power supply unit is activated, the supply of power to the CPU or the like is not delayed in the meantime. Countermeasures are required. In this case, during the period from when the power supply unit is started until the supplied power is stable and in the operating state (hereinafter referred to as “stable waiting time”), the CPU is set in the power saving state and the stable waiting time has elapsed. The CPU is returned to the normal state.

JP 2006-302059 A

  However, if the accuracy of predicting CPU power consumption is low in the first place, it will be difficult to accurately determine the number of power supply units to be operated. The fact that the number of power supply units to be operated cannot be determined with high accuracy means that, as an example, as shown in FIG. 8, although one power supply unit can actually cover power consumption, There is a risk of causing the system to operate, and in this case, power is wasted. On the contrary, if the power consumption of the CPU suddenly increases during the stable waiting time, as shown in FIG. 9 as an example, there is a possibility of causing a situation where the supply of power is not in time for the power consumption of the CPU.

In one aspect, an object of the present invention is to accurately estimate the time when the power supplied from the power supply unit reaches the maximum power that can be supplied from the power supply unit.
In another aspect, an object of the present invention is to perform start-up control of a power supply in consideration of a time required until the activated power supply is in an operating state in which power can be supplied.

  In the disclosed technology, power can be supplied to the information processing unit by a plurality of power supply units, and power supplied from the operating power supply unit to the information processing unit among the plurality of power supply units by the first measurement unit. The corresponding physical quantity is measured at different time points. In the disclosed technology, the estimation unit supplies the power supplied from the operating power supply unit to the information processing unit by the operating power supply unit based on the measurement result history of the first measurement unit. Estimate the time to reach a predetermined threshold for possible power.

  According to one embodiment, there is an effect that it is possible to accurately estimate the time when the power supplied from the power supply unit reaches the maximum power that can be supplied from the power supply unit.

It is a functional block diagram which shows an example of the principal part function of the server apparatus which concerns on embodiment. It is a block diagram which shows an example of the principal part structure of the electric system of the server apparatus which concerns on embodiment. It is a schematic diagram which shows an example of a structure of maximum supplyable electric power DB which concerns on embodiment. It is a schematic diagram which shows an example of a structure of stable waiting time DB which concerns on embodiment. It is a flowchart which shows an example of the flow of the electric power feeding control process which concerns on embodiment. It is a graph which shows an example of the approximate line created by the estimation part which concerns on embodiment. It is a graph which shows an example of transition of supply electric power when changing a CPU from a non-power-saving state to a power-saving state, and returning to a non-power-saving state again by a control part concerning an embodiment. It is a graph which shows an example of the relationship between the supplied power and the actual power consumption when the power is consumed wastefully in the conventional server device. It is a graph which shows an example of the relationship between the power supply when the power supply capability in the conventional server apparatus is insufficient, and actual power consumption.

  Hereinafter, an example of an embodiment of the disclosed technology will be described in detail with reference to the drawings. In the following description, as an example of the information processing apparatus according to the disclosed technology, driving power is supplied from a plurality of power supply units to a plurality of system boards each including a processing unit including a CPU and a memory. A server apparatus having a configuration as described above will be described as an example. However, the disclosed technique is not limited to this. The disclosed technology can also be applied to various information processing apparatuses such as a stationary personal computer and a parallel arithmetic device.

  FIG. 1 shows an example of main functions of the server device 10 according to the present embodiment. The server device 10 includes a plurality of system boards 12, a plurality of power supply units 14, and a service processor 16 that are examples of information processing units. The system board 12 includes a CPU 18 that is an example of a processing unit, a memory 22, and a current sensor 24 that is an example of a second measurement unit according to the disclosed technology. The CPU 18 executes a predetermined process on the input information. Further, the CPU 18 has a power saving function and can be operated in a non-power saving state and a power saving state. The CPU 18 normally operates in the non-power-saving state, and switches from the operation in the non-power-saving state to the operation in the power-saving state by using the power-saving function. Then, the operation of the power saving function is canceled to return from the operation in the power saving state to the operation in the non-power saving state. In addition, examples of the “power saving state” mentioned here include an operation stop state such as a so-called sleep state and a standby state, and a low operation state that consumes less power than normal operation. (State in which the number of operation clocks per unit time is smaller than in normal operation). Therefore, the “non-power-saving state” mentioned here means an operating state that consumes more power than the “power-saving state”.

  The memory 22 includes a nonvolatile storage unit and a volatile storage unit. An example of a nonvolatile storage unit is a flash memory. In the nonvolatile storage unit, various programs executed by the CPU 18 and various parameters used by the CPU 18 are stored in advance. An example of a volatile storage unit is a RAM (Random Access Memory). The volatile storage unit is used as a work area when the CPU 18 executes various programs.

  The current sensor 24 is provided corresponding to each of the CPUs 18 mounted for each system board 12 and measures a physical quantity corresponding to the power consumption of the corresponding CPU 18. In the present embodiment, the magnitude of the current supplied to the CPU 18 (hereinafter referred to as “current value”) is adopted as an example of “physical quantity”. The “physical quantity” is not limited to the current value, but may be the magnitude (voltage value) of the voltage supplied to the CPU 18.

  The plurality of power supply units 14 supply power for operating the plurality of system boards 12 to the plurality of system boards 12. In the present embodiment, for example, the plurality of power supply units 14 are connected to the plurality of system boards 12 via the power supply lines 26, and convert the commercial alternating current (AC) supplied from the outlets into direct currents to convert the power supply lines. The power is supplied to the plurality of system boards 12 via 26. Note that the power supplied to the system board 12 is mainly consumed by the CPU 18.

  Each power supply unit 14 includes a current sensor 28 which is an example of a first measurement unit according to the disclosed technique. The current sensor 28 measures a physical quantity corresponding to power supplied from the power supply unit 14 to the plurality of system boards 12 every predetermined time. Further, the “predetermined time” referred to here includes, for example, a time obtained by multiplying a stable waiting time of the power supply unit 14 by a predetermined coefficient that is a positive value less than 1. For example, when the stabilization waiting time of the power supply unit 14 is 0.5 seconds and the predetermined coefficient is “0.01”, the predetermined time is “0.005 seconds”. Further, instead of measuring at “every predetermined time”, the measurement may be performed at a plurality of different time points before the estimation process by the estimation unit 32 described later.

  The service processor 16 controls the overall operation of the server device 10. The service processor 16 is connected to a plurality of system boards 12 and a plurality of power supply units 14 via a bus 30 including an address bus and a system bus. Accordingly, the service processor 16 can acquire the current values measured by the current sensors 24 and 28, control each of the plurality of system boards 12, and control each of the plurality of power supply units 14.

  The service processor 16 includes an estimation unit 32 and a control unit 34. Based on the history of measurement results from the current sensor 28, the estimation unit 32 reaches the predetermined threshold regarding the power that can be supplied from the operating power supply unit 14 from the operating power supply unit 14 to the system board 12. Estimate the time until. Here, as an example, the estimation unit 32 estimates the time when the maximum power that can be supplied by the operating power supply unit 14 is reached (hereinafter referred to as “supplyable power arrival time”). However, the disclosed technique is not limited to this. For example, the time when the maximum power that can be supplied reaches a value that includes some margin is estimated. However, the disclosed technique is not limited to this. For example, it is possible to estimate the time to reach a value with some margin added to the maximum supplyable power (time to reach a value with some margin added to the maximum supplyable power). For example, the estimation unit 32 creates an approximate line related to the time change of the current value measured at predetermined times of the current sensor 28 and estimates the supplyable power arrival time based on the created approximate line. Although an approximate straight line is employed here, the disclosed technique is not limited to this, and may be an approximate curve, for example.

  When the non-operating power supply unit 14 is activated, the control unit 34 performs control to activate the non-operating power supply unit 14 and start supplying power when a predetermined condition is satisfied. Here, “when the predetermined condition is satisfied” means that, for example, the time required to reach a predetermined threshold estimated by the estimation unit 32 drives the non-operating power supply unit 14 and the power supply unit 14 can supply power. This is a case where it is the same as or longer than the time required to reach a proper operating state. That is, the supplyable power arrival time is the same as or later than the time until the power supply unit 14 enters an operating state in which power can be supplied. That is, when it is predicted (estimated) that the power supplied to the system board 12 will reach the maximum power that can be supplied by the power supply unit 14 operating in the near future based on the history of the measurement results of the current sensor 28. The unit 34 activates the non-operating power supply unit 14 to start power supply. The control unit 34 uses the estimation result of the estimation unit 32 to determine whether or not the additional power supply unit 14 needs to be turned on, and when the power supply unit 14 needs to be turned on, the control unit 34 sets the CPU 18 to the power saving state. Make a decision as to whether or not it is necessary.

  When the supplyable power arrival time is earlier than the time after the stabilization waiting time from the current time, the control unit 34 is in a non-power-saving state among the plurality of CPUs 18 until the activated power supply unit 14 is in an operating state. The operating state of the CPU 18 that is operating and satisfies a predetermined condition is switched to a power saving state. In the present embodiment, as an example of the “predetermined condition”, the condition that the current value measured by the plurality of current sensors 24 is the maximum value is employed. However, the disclosed technique is not limited to this. For example, another example of the “predetermined condition” is a condition that the current value measured by the corresponding current sensor 24 exceeds a predetermined threshold value. Here, the “predetermined threshold value” is an upper limit value of a current value corresponding to power consumption that does not have insufficient power supply capability, for example, and refers to a value obtained in advance by an experiment or simulation using an actual machine. .

  In addition, the control unit 34 has a plurality of operating power supply units 14 and the total power supplied to the plurality of system boards 12 specified by the current values measured by the plurality of current sensors 28 reaches a predetermined power. If not, the operation of one power supply unit 14 in operation is stopped. Here, “the case where the predetermined power is not reached” means, for example, the maximum suppliable power of the power supply unit 14 remaining in operation when the operation of one power supply unit 14 in operation is stopped. It refers to the case where it has not been reached.

  FIG. 2 shows an example of the main configuration of the electrical system of the service processor 16. As shown in FIG. 2, the service processor 16 includes a CPU 42, a memory 44, and a nonvolatile storage unit 46, which are connected to each other via the bus 30. The storage unit 46 can be realized by an HDD (Hard Disk Drive), a flash memory, or the like. The storage unit 46 serving as a recording medium stores a power supply control processing program 50, a suppliable maximum power DB (database) 52, and a stable waiting time DB 54 (details will be described later).

  The CPU 42 reads the power supply control processing program 50 from the storage unit 46 and expands it in the memory 44, and sequentially executes the processes of the power supply control processing program 50. The power supply control processing program 50 includes an estimation process 56 and a control process 58. The CPU 42 operates as the estimation unit 32 illustrated in FIG. 1 by executing the estimation process 56. The CPU 42 operates as the control unit 34 illustrated in FIG. 1 by executing the control process 58.

  In addition, although the case where the power feeding control processing program 50 is read from the storage unit 46 is illustrated here, it is not necessarily stored in the storage unit 46 from the beginning. For example, a power supply control processing program 50 is first applied to an arbitrary “portable physical medium” such as a flexible disk connected to the service processor 16, a so-called FD, CD-ROM, DVD disk, magneto-optical disk, IC card or the like. May be stored. Then, the service processor 16 may acquire and execute the power supply control processing program 50 from these portable physical media. Further, the power supply control processing program 50 may be stored in another computer or server device connected to the service processor 16 via the Internet or a LAN (Local Area Network). In this case, for example, a mode in which the service processor 16 acquires and executes the power supply control processing program 50 from these can be exemplified.

  FIG. 3 schematically shows an example of the configuration of the maximum power supply DB 52 that can be supplied. As shown in FIG. 3, the maximum supplyable power DB 52 stores a power supply unit ID (identification) and a maximum supplyable power value in a one-to-one correspondence. The “power supply unit ID” referred to here is, for example, an ID given in advance to each of the plurality of power supply units 14 and uniquely identifying each power supply unit 14. The “maximum suppliable power value” mentioned here refers to the maximum power value that can be supplied to the plurality of system boards 52 by the power supply unit 14 specified by the power supply unit ID, for example. Therefore, the CPU 42 can uniquely derive the value of the maximum power that can be supplied from the power supply unit ID with reference to the maximum power DB 52 that can be supplied.

  FIG. 4 schematically shows an example of the configuration of the stable waiting time DB 54. As shown in FIG. 4, the power supply unit ID and the stable waiting time are stored in the stable waiting time DB 54 in a one-to-one correspondence. That is, for each power supply unit 14 specified by the power supply unit ID, the corresponding waiting time for the power supply unit 14 is associated. Therefore, the CPU 42 can uniquely derive the stable waiting time from the power supply unit ID with reference to the stable waiting time DB 54.

  Next, as an operation of the present embodiment, a power supply control process performed by the service processor 16 of the server apparatus 10 when the CPU 42 executes the power supply control process program 50 will be described with reference to FIG. FIG. 5 is a diagram illustrating an example of the power supply control process in the case where the power supply unit 14 that supplies power to the CPU 42 is added and activated due to an increase in the processing load of the CPU 42 or the like. First, in step 100, the control unit 34 acquires the value of the maximum power that can be supplied from one power supply unit 14 that is the activation target from the maximum supplyable power DB 52. In step 100, if there is a power supply unit 14 that is already in operation, the acquired value of the maximum suppliable power is added to the maximum suppliable power of the power supply unit 14 that is already in operation, The maximum power that can be supplied by the power supply unit 14 is calculated.

  In the next step 102, the control unit 34 activates one power supply unit 14 to be activated, and acquires a stable waiting time associated with the activated power supply unit 14. And the control part 34 calculates the stable operation time (time after the acquired stable waiting time from the present time) when the activated power supply part 14 will be in the operation state in which supplied power is stabilized. In the next step 104, the estimation unit 32 obtains the current value measured by the current sensor 28 in each of the plurality of power supply units 14. In the next step 106, the estimation unit 32 stores the current value for each current sensor 28 acquired in step 104 in the memory 44 in association with the current time.

  In the next step 108, the estimation unit 32 determines whether or not the current value has been acquired a predetermined number of times in the above step 104. That is, it is determined whether or not the above step 104 has been performed a predetermined number of times (for example, three times). In step 108, if the current value has not been acquired a predetermined number of times in step 104, the determination is negative and the routine proceeds to step 112. In step 112, the estimation unit 32 determines whether or not a predetermined time (for example, 0.005 seconds) has elapsed since the above step 104 was performed (since the latest acquisition of the current value). In step 112, if a predetermined time has elapsed since step 104 was performed, the determination is affirmed and the routine proceeds to step 104. In step 112, if the predetermined time has not elapsed since step 104 is performed, the determination is denied and the determination in step 112 is performed again.

  On the other hand, if the current value is acquired a predetermined number of times in the above step 104 in step 108, the determination is affirmed and the routine proceeds to step 110. In step 110, the estimation unit 32 estimates the supplyable power arrival time for the power supply unit 14 activated in step 102. That is, in this step 110, the estimation unit 32 estimates the supplyable power arrival time based on the current value history for each predetermined time for each current sensor 28 stored in the memory 44 in the above step 106. For example, first, the sum of the supplied power for each predetermined time is calculated from the current value for each predetermined time for each current sensor 28 stored in the memory 44, and an approximate straight line regarding the time change of the calculated sum of the supplied power is created. The The supplyable power arrival time is estimated by specifying the supplyable power arrival time from the generated approximate line and the value of the maximum supplyable power acquired in step 100 above. For example, as shown in FIG. 6, the sum of the supplied power is calculated for each of the times T1, T2, and T3, and an approximate straight line based on, for example, the least square method is created for the calculated supplied power. And the time corresponding to the intersection of the created approximate straight line and the straight line indicating the maximum power that can be supplied is specified as the supplyable power arrival time. When the approximate straight line has a slope of “0” or less, the previously estimated supplyable power arrival time is adopted, and when the previous estimated supplyable power arrival time does not exist, a predetermined slope (default setting) is used. The slope of the positive value) is adopted as the slope of the approximate line. If the slope of the approximate line is “0” or less, the supplyable power arrival time may be set to ∞ (infinite) for convenience.

  In the next step 111, the suppliable power arrival time estimated in step 110 is stored in the memory 44 by the estimation unit 32. In the next step 114, the control unit 34 determines whether or not the latest supplyable power arrival time stored in the memory 44 is earlier than the stable operation waiting time calculated in step 102. In this step 114, when the latest supplyable power arrival time is earlier than the stable operation waiting time, the determination is affirmed and the routine proceeds to step 116.

  When the determination in step 114 is affirmed, there is a possibility that the supply power is insufficient before the time when the supply power of the power supply unit 14 activated in the previous step 102 becomes stable. Therefore, in step 116, the control unit 34 acquires the current value measured by the current sensor 24 in each of the plurality of system boards 12. In the next step 118, the control unit 34 specifies the CPU 18 to be set in the power saving state from the CPUs 18 in the non power saving state based on the current value acquired in step 116. That is, the CPU 18 having the maximum value among the current values acquired in step 116 is specified as a target to be set in the power saving state. Note that the disclosed technology is not limited to this. For example, in step 118, the control unit 34 specifies one or more CPUs 18 having a current value that exceeds the predetermined threshold value in the current value acquired in step 116 as a target to be set in the power saving state. Also good.

  In the next step 120, the control unit 34 puts the CPU 18 specified in the above step 118 into a power saving state. That is, here, the CPU 18 having the maximum value among the current values acquired in the above step 116 is set in the power saving state. Therefore, the power consumption can be greatly reduced as compared with the case where the CPU 18 having the maximum value among the current values acquired in step 116 is not set in the power saving state. Note that instead of the CPU 18 having the maximum value among the current values acquired in the above step 116, the CPU 18 having the minimum value among the current values acquired in the above step 116, or a plurality of the power values in ascending order of power consumption. The CPU 18 may be in a power saving state. In this case, as compared with the case where the CPU 18 having the maximum current value acquired in step 116 is set in the power saving state, the processing by the CPU 18 operating at a high load is suppressed from being adversely affected. can do.

  In step 122, the control unit 34 determines whether or not the stable operation time has come. In this step 122, when the stable operation time has not come, the determination is denied and the determination of this step 122 is performed again. If the stable operation time has come in step 122, the determination is affirmed and the routine proceeds to step 124. In step 124, the control unit 34 causes the CPU 18, which has been in the power saving state in step 122, to return to the non-power saving state, and then proceeds to step 134.

  As described above, when it is predicted that the latest supplyable power arrival time will arrive before the stable operation time arrives (when the determination in step 114 is affirmative), as shown in FIG. The CPU 18 in the power state is in a power saving state (step 120). Thereby, a shortage of power supply can be avoided in advance. Then, when the stable operation time arrives, the CPU 18 in the power saving state is returned to the non-power saving state (step 124), thereby avoiding the state where the processing capability of the CPU 18 is unnecessarily suppressed. Can do.

  On the other hand, if the latest supplyable power arrival time is later than the stable operation time in step 114, the determination is negative and the routine proceeds to step 125. In step 125, the control unit 34 determines whether or not the stable operation time has come. In step 125, when the stable operation time has not come, the determination is denied and the determination in step 125 is performed again. If the stable operation time has come in step 125, the determination is affirmed and the routine proceeds to step 126. In step 126, the control unit 34 determines whether or not a time before the stable waiting time of the next power supply unit 14 to be activated has arrived before the latest supplyable power arrival time. In this step 126, when the time before the stable waiting time of the next power supply unit 14 to be activated comes after the latest supplyable power arrival time, the determination is affirmed and the routine proceeds to step 132. In this case, if there is an inactive power supply unit 14, the power supply unit 14 is activated. In this step 126, when the time before the stable waiting time of the next power supply unit 14 to be activated has not arrived before the latest supplyable power arrival time, the determination is negative and the routine proceeds to step 128.

  In step 128, the control unit 34 determines whether or not the supplyable power arrival time to be compared is currently stored in the memory 44. That is, it is determined whether or not a plurality of supplyable power arrival times are stored in the memory 44. In step 128, when the supplyable power arrival time to be compared is not stored in the memory 44, the determination is negative and the process proceeds to step 104. In this case, a new supplyable power arrival time is estimated. In step 128, when the supplyable power arrival time to be compared is stored in the memory 44, the determination is affirmed and the process proceeds to step 130.

  In step 130, the control unit 34 determines whether or not the latest supplyable power arrival time stored in the memory 44 is earlier than the previous supplyable power arrival time stored in the memory 44. In this step 130, when the latest supplyable power arrival time stored in the memory 44 is earlier than the previous supplyable power arrival time stored in the memory 44, the determination is affirmed and the process proceeds to step 132.

  If the determination in step 130 is affirmed, it can be determined that the increase rate of the power consumption of the system board 12 tends to increase and there is a possibility that the shortage of the supplied power may occur early. Therefore, in step 132, the control unit 34 determines whether all the power supply units 14 are currently activated. If all the power supply units 14 are not activated in step 132, the determination is denied and the process proceeds to step 100. In this case, the non-operating power supply unit 14 is activated. If all the power supply units 14 are activated in step 132, the determination is affirmed and the routine proceeds to step 134. In the present embodiment, the suppliable power of each power supply unit 14 and the number of power supply units 14 are determined according to the maximum power consumption of the entire server device 10. Therefore, when the determination in step 132 is affirmed, there is no shortage of supply power.

  In step 134, the control unit 34 determines whether or not an end condition of the power supply control process (for example, an instruction to stop the operation of the server device 10 is input to the CPU 42) is satisfied. In step 134, when the end condition of the power supply control process is satisfied, the determination is affirmed and the power supply control process ends. In step 134, if the end condition of the power supply control process is not satisfied, the determination is negative and the process proceeds to step 104.

  On the other hand, in step 130, if the latest supplyable power arrival time stored in the memory 44 is not earlier than the previous supplyable power arrival time stored in the memory 44, the determination is negative and the process proceeds to step 136.

  If the determination in step 130 is negative, the system board 12 can determine that although the power consumption has increased, the increase rate tends to decrease or the power consumption tends to decrease. Therefore, in step 138, it is determined whether or not the latest supplyable power arrival time is a predetermined period or more. In step 138, if the latest supplyable power arrival time is not more than a predetermined period, the determination is affirmed and the routine proceeds to step 104. In step 138, if the latest supplyable power arrival time is a predetermined period or more, the determination is affirmed and the routine proceeds to step 140.

  In step 140, the control unit 34 determines whether or not power supply is not insufficient even if the operation of one of the power supply units 14 currently operating is stopped. In step 140, if the operation of one power supply unit 14 is stopped and power supply is insufficient, the determination is denied and the process proceeds to step 104. In step 140, if the power supply is not insufficient even if the operation of one power supply unit 14 is stopped, the determination is affirmed and the process proceeds to step 142. In step 142, the control unit 24 stops the operation of one of the power supply units 14 currently operating, and then the process proceeds to step 104.

  As described above in detail, according to the server device 10 according to the present embodiment, the suppliable power arrival time is periodically estimated and updated by the estimation unit 32. For this reason, when it is expected that the supplyable power arrival time will arrive before the stable time of supply power comes, the occurrence of a situation where the power supply capability is insufficient is suppressed by setting the CPU 18 in the power saving state. can do. In addition, when the stable operation time has arrived, it is possible to suppress the occurrence of a situation where the power supply capability is insufficient by additionally operating the power supply unit 14. In addition, even when the most recently acquired supplyable power arrival time is expected to arrive earlier than the supplyable power arrival time acquired in the past, the power supply capability can be increased by additionally operating the power supply unit 14. The occurrence of a shortage can be suppressed. Furthermore, when the latest available power supply arrival time is expected to come after the supply power arrival time acquired in the past and various conditions are satisfied, the power supply unit 14 is operated. Since it has stopped, the unnecessary consumption of electric power can be suppressed.

  In addition, although the aspect which calculates the power consumption of CPU18 by measuring the electric current value supplied to CPU18 with the current sensor 24 was illustrated above, the technique of an indication is not limited to this. For example, the operation clock frequency of the CPU 18 may be detected or acquired, and the power consumption of the CPU 18 may be calculated from the detected or acquired operation clock frequency. The current supplied to the CPU 18 is the sum of the dynamic current and the leakage current. Of these, the dynamic current changes according to the operation clock frequency, and the leakage current changes depending on the voltage and temperature. Therefore, when calculating the dynamic current from the operation clock frequency and calculating the power consumption, the leakage current may be regarded as a constant value, or the leakage current may be calculated by detecting or acquiring the voltage and temperature. Further, for example, the power consumption of the CPU 18 may be estimated from the processing load of the program executed by the CPU 18.

  In the above description, the current sensor 24 is provided for each power supply unit 14, but the disclosed technique is not limited to this, and a configuration for measuring the current value of the power supply line 26 may be adopted. . In this case, it is preferable to provide a single current sensor 24 for measuring the current value of the power supply line 26 to which the currents output from the plurality of power supply units 14 have joined, because the configuration becomes simple.

  Moreover, although the server apparatus 10 illustrated the aspect which has the some system board 12 above, the technique of an indication is not limited to this, and it is materialized even if the system board 12 is single. Needless to say.

  All documents, patent applications and technical standards mentioned in this specification are to the same extent as if each individual document, patent application and technical standard were specifically and individually stated to be incorporated by reference. Incorporated by reference in the book.

Claims (22)

A plurality of power supply units capable of supplying power to the information processing unit;
A first measurement unit that measures a physical quantity according to power supplied from an operating power source unit to the information processing unit among the plurality of power source units at different time points;
Based on the history of measurement results by the first measurement unit, the supply power from the operating power supply unit to the information processing unit reaches a predetermined threshold related to the power that can be supplied by the operating power supply unit. An estimation unit for estimating time;
An information processing apparatus including:   The estimation unit creates an approximate line related to temporal changes in the physical quantity measured at the plurality of different time points by the first measurement unit, and estimates a time until the predetermined threshold is reached based on the created approximate line The information processing apparatus according to claim 1.   When starting the non-operating power supply unit, the time until reaching the predetermined threshold estimated by the estimation unit is the operating state in which the non-operating power supply unit is started and the power supply unit can supply power The control part which further controls to start the supply of electric power by starting the said non-operating power supply part when it is the same as the required time until becoming or longer than this required time. The information processing apparatus described in 1. The information processing unit includes a plurality of processing units operable in a non-power saving state and a power saving state, respectively.
When the time until the control unit reaches a predetermined threshold estimated by the estimation unit is shorter than the required time, the control unit is configured to wait until the activated power supply unit is in an operating state. The information processing apparatus according to claim 3, wherein an operation state of a processing unit that operates in the non-power-saving state and satisfies a predetermined condition is switched to the power-saving state. The information processing unit includes a plurality of processing units operable in a non-power saving state and a power saving state, respectively.
When the time to reach the predetermined threshold estimated by the estimation unit is shorter than the required time until the power supply unit that is not operating is activated and the power supply unit is in an operating state in which power can be supplied, The operating state of the processing unit that is operating in the non-power-saving state and satisfies a predetermined condition among the plurality of processing units until the activated power supply unit is in an operating state is the saving state. The information processing apparatus according to claim 1, further comprising a control unit that switches to a power state. A plurality of second measuring units provided corresponding to each of the plurality of processing units and measuring a physical quantity corresponding to the power consumption of each of the plurality of processing units;
The information processing apparatus according to claim 4, wherein the predetermined condition is a condition that the physical quantity measured by the corresponding second measurement unit exceeds a predetermined threshold. A plurality of second measuring units provided corresponding to each of the plurality of processing units and measuring a physical quantity corresponding to power consumption for each of the plurality of processing units;
The information processing apparatus according to claim 4, wherein the predetermined condition is a condition that a maximum value among the physical quantities measured by the plurality of second measurement units. A plurality of power supply units capable of supplying power to the information processing unit;
An estimation unit that estimates a time until power consumed by the information processing unit reaches a predetermined threshold;
When starting the non-operating power supply unit, the time until the predetermined threshold estimated by the estimation unit reaches the non-operating power supply unit is activated, and the power supply unit can supply power A control unit that controls to start the power supply by starting the power supply unit that is not in operation when the required time to be equal to or longer than the required time;
An information processing apparatus including: A supply step of supplying power to the information processing unit by a plurality of power supply units;
A first measurement step of periodically measuring a physical quantity according to power supplied from an operating power supply unit to the information processing unit among the plurality of power supply units;
Based on the history of measurement results in the first measurement step, the power supplied from the operating power supply unit to the information processing unit reaches a predetermined threshold regarding the power that can be supplied by the operating power supply unit. An estimation step for estimating time;
Including a power supply control method.   The estimating step creates an approximate line related to the time change of the physical quantity measured periodically by the first measurement step, and estimates a time until the predetermined threshold is reached based on the created approximate line. Item 10. The power feeding control method according to Item 9.   When starting the non-operating power supply unit, the time until the predetermined threshold estimated by the estimation step is reached is the operating state where the non-operating power supply unit is started and the power supply unit can supply power 11. The control step of controlling to start the supply of electric power by starting the non-operating power supply unit when it is equal to or longer than the required time until the power consumption is reached. The power feeding control method described in 1. The information processing unit includes a plurality of processing units operable in a non-power saving state and a power saving state, respectively.
In the control step, when the time until the predetermined threshold estimated in the estimation step is shorter than the required time, the plurality of processing units are in a period until the activated power supply unit is in an operating state. The power supply control method according to claim 11, wherein an operation state of a processing unit that operates in the non-power-saving state and satisfies a predetermined condition is switched to the power-saving state. The information processing unit includes a plurality of processing units operable in a non-power saving state and a power saving state, respectively.
When the time to reach the predetermined threshold estimated in the estimation step is shorter than the required time until the power supply unit that is not operating is activated and the power supply unit is in an operating state capable of supplying power, The operating state of the processing unit that is operating in the non-power-saving state and satisfies a predetermined condition among the plurality of processing units until the activated power supply unit is in an operating state is the saving state. The power supply control method according to claim 9 or 10, further comprising a control step of switching to a power state. A second measuring step of measuring a physical quantity corresponding to power consumption of each of the plurality of processing units by each of a plurality of second measuring units provided corresponding to each of the plurality of processing units;
The power supply control method according to claim 12 or 13, wherein the predetermined condition is a condition that the physical quantity measured by the corresponding second measurement step exceeds a predetermined threshold. The method further includes a second measurement step of measuring a physical quantity corresponding to power consumption for each of the plurality of processing units by each of the plurality of second measurement units provided corresponding to each of the plurality of processing units. ,
The power supply control method according to claim 12 or 13, wherein the predetermined condition is a condition that a maximum value among the physical quantities measured by the plurality of second measuring units. On the computer,
A supply step of supplying power to the information processing unit by a plurality of power supply units;
A first measurement step of periodically measuring a physical quantity according to power supplied from an operating power supply unit to the information processing unit among the plurality of power supply units;
Based on the history of measurement results in the first measurement step, the power supplied from the operating power supply unit to the information processing unit reaches a predetermined threshold regarding the power that can be supplied by the operating power supply unit. An estimation step for estimating time;
The power supply control program for performing the process containing.   The estimating step creates an approximate line related to the time change of the physical quantity measured periodically by the first measurement step, and estimates a time until the predetermined threshold is reached based on the created approximate line. Item 17. The power supply control program according to Item 16.   When starting the non-operating power supply unit, the time until the predetermined threshold estimated by the estimation step is reached is the operating state where the non-operating power supply unit is started and the power supply unit can supply power The control step of controlling to start the supply of electric power by starting the non-operating power supply unit when it is equal to or longer than the required time until the power consumption is reached. The power supply control program described in 1. The information processing unit includes a plurality of processing units operable in a non-power saving state and a power saving state, respectively.
In the control step, when the time until the predetermined threshold estimated in the estimation step is shorter than the required time, the plurality of processing units are in a period until the activated power supply unit is in an operating state. The power supply control program according to claim 18, wherein an operation state of a processing unit that operates in the non-power-saving state and satisfies a predetermined condition is switched to the power-saving state. The information processing unit includes a plurality of processing units operable in a non-power saving state and a power saving state, respectively.
When the time to reach the predetermined threshold estimated in the estimation step is shorter than the required time until the power supply unit that is not operating is activated and the power supply unit is in an operating state capable of supplying power, The operating state of the processing unit that is operating in the non-power-saving state and satisfies a predetermined condition among the plurality of processing units until the activated power supply unit is in an operating state is the saving state. The power supply control program according to claim 17 or 18, further comprising a control step of switching to a power state. A second measuring step of measuring a physical quantity corresponding to power consumption of each of the plurality of processing units by each of a plurality of second measuring units provided corresponding to each of the plurality of processing units;
The power supply control program according to claim 19 or 20, wherein the predetermined condition is a condition that the physical quantity measured by the corresponding second measurement step exceeds a predetermined threshold. The method further includes a second measurement step of measuring a physical quantity corresponding to power consumption for each of the plurality of processing units by each of the plurality of second measurement units provided corresponding to each of the plurality of processing units. ,
The power supply control program according to claim 19 or 20, wherein the predetermined condition is a condition that the maximum value among the physical quantities measured by the plurality of second measurement units.