Adaptive control circuit, adaptive control method and chip for chip power consumption

Abstract

The present application discloses an adaptive control circuit, an adaptive control method and a chip for chip power consumption, and belongs to the field of chip technology. The temperature acquisition module is used to send the minimum temperature parameter of the chip to the dynamic compensation module; the process acquisition module is used to send the minimum process parameter of the chip to the dynamic compensation module; the dynamic compensation module is used to generate a simulation formula for the minimum power supply voltage surface of the chip, and the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter are input into the simulation formula, and the calculation result is sent to the voltage feedback module as the target resistance value; the voltage feedback module is used to adjust the resistance value of the variable resistor to the target resistance value, adjust the current feedback voltage to the target feedback voltage according to the adjusted variable resistance, and send the current feedback voltage to the power supply module; the power supply module is used to adjust the power supply voltage of the chip according to the current feedback voltage. The present application can reduce the power consumption of the chip and improve the consistency and profit margin of the system.

Description

Technical Field

[0001] The present application relates to the field of chip technology, and in particular to an adaptive control circuit, an adaptive control method and a chip for chip power consumption. Background Art

[0002] As chip manufacturing processes evolve to the deep submicron era, chips present the following two important characteristics: (1) The static power consumption Ps of the chip shows an exponential upward trend with the increase of temperature; (2) The static power consumption Ps of the chip shows greater differences with the deviation of the process. The static power consumption Ps refers to the basic power consumption consumed by the chip when it is not performing any switching activities or working state.

[0003] To deal with the increase in power consumption caused by temperature rise, we can divide the operating temperature into several continuous temperature intervals within the full temperature range and use different supply voltages in different temperature intervals. However, simply adjusting the supply voltage according to the operating temperature has limited power consumption optimization, and the temperature may cross to adjacent temperature intervals during voltage switching. Adjusting the supply voltage needs to take into account the voltage requirements of adjacent temperature intervals, which reduces the range of supply voltage adjustment and thus reduces the power consumption optimization.

[0004] For the increase in power consumption caused by process deviation, we can classify the chips by model according to the process differences, ship the chips according to different model categories, and set different power supply voltages for different chip models to optimize power consumption. However, classification according to process differences will increase the number of chip models, which brings huge challenges to production control and capacity management. In addition, since the number of chip models cannot be increased indefinitely, there will be certain deviations in the chip process under the same model. In order to meet the power supply requirements of the chip with the largest deviation among the same model, the power supply requirements of the chips under the same model are generally high, and the power consumption optimization effect is affected. Summary of the invention

[0005] The present application provides an adaptive control circuit, adaptive control method and chip for chip power consumption, which are used to solve the problem that the power supply voltage is controlled according to the divided temperature range or model, so that the power consumption optimization range of the chip is limited. The technical solution is as follows:

[0006] According to a first aspect of the present application, there is provided an adaptive control circuit for chip power consumption, the adaptive control circuit comprising a temperature acquisition module, a process acquisition module, a dynamic compensation module, a voltage feedback module and a power supply module;

[0007] The temperature acquisition module is used to acquire temperature parameters of various areas in the chip and send the minimum temperature parameter to the dynamic compensation module;

[0008] The process acquisition module is used to collect process parameters of each area in the chip and send the minimum process parameter to the dynamic compensation module;

[0009] The dynamic compensation module is used to generate a simulation formula of the minimum supply voltage surface of the chip, input the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and send the calculation result of the simulation formula as the target resistance value to the voltage feedback module, and the minimum supply voltage surface is simulated according to the temperature extreme value and process extreme value of the chip;

[0010] The voltage feedback module is used to adjust the resistance value of the variable resistor to the target resistance value, adjust the current feedback voltage to the target feedback voltage according to the adjusted variable resistor, and send the current feedback voltage to the power module;

[0011] The power supply module is used to adjust the power supply voltage of the chip according to the current feedback voltage.

[0012] In a possible implementation, the calibration parameters include an initial minimum voltage parameter, a temperature gain parameter, a temperature offset parameter, a process gain parameter, and a process offset parameter;

[0013] In the simulation formula, the supply voltage is equal to the initial minimum voltage parameter minus the temperature adjustment parameter and the process adjustment parameter, the temperature adjustment parameter is equal to the sum of the minimum temperature parameter and the temperature offset parameter multiplied by the temperature gain parameter, and the process adjustment parameter is equal to the sum of the minimum process parameter and the process offset parameter multiplied by the process gain parameter.

[0014] In a possible implementation, the dynamic compensation module includes at least one of a temperature protection unit, a process protection unit, and a resistance protection unit;

[0015] The temperature protection unit is used to obtain a maximum temperature adjustment parameter and a minimum temperature adjustment parameter in the calibration parameters, and limit the temperature adjustment parameter to between the maximum temperature adjustment parameter and the minimum temperature adjustment parameter;

[0016] The process protection unit is used to obtain a maximum process adjustment parameter and a minimum process adjustment parameter in the calibration parameters, and limit the process adjustment parameter to between the maximum process adjustment parameter and the minimum process adjustment parameter;

[0017] The resistance protection unit is used to obtain a maximum resistance value and a minimum resistance value in the calibration parameters, and to limit the target resistance value between the maximum resistance value and the minimum resistance value.

[0018] In a possible implementation, the dynamic compensation module further includes a temperature validity checking unit;

[0019] The temperature validity checking unit is used to detect whether the minimum temperature parameter is within the normal operating temperature range;

[0020] If the minimum temperature parameter is not within the normal operating temperature range, an error signal is generated and sent to the temperature protection unit;

[0021] The temperature protection unit is used to adjust the temperature gain parameter according to the error signal to turn off the temperature adjustment function.

[0022] In a possible implementation, the dynamic compensation module further includes a process validity checking unit;

[0023] The process validity checking unit is used to detect whether the difference between the minimum process parameter and the initial process parameter exceeds a threshold value;

[0024] If the difference between the minimum process parameter and the initial process parameter exceeds a threshold, an error signal is generated and sent to the process protection unit;

[0025] The process protection unit is used to adjust the process gain parameter according to the error signal to turn off the process adjustment function.

[0026] In a possible implementation, the adaptive control circuit further includes a fault detection module, the temperature acquisition module includes a plurality of temperature sensors and an extreme temperature extraction unit, and the plurality of temperature sensors are distributed in various areas of the chip;

[0027] Each temperature sensor is used to collect temperature parameters of a corresponding area and send the temperature parameters to the extreme temperature extraction unit;

[0028] The extreme temperature extraction unit is used to extract a maximum temperature parameter and a minimum temperature parameter from a plurality of temperature parameters, send the maximum temperature parameter to the fault detection module, and send the minimum temperature parameter to the dynamic compensation module;

[0029] The fault detection module is used to detect whether to issue an over-temperature warning according to the maximum temperature parameter.

[0030] In a possible implementation, the process acquisition module includes a plurality of process sensors and an extreme value process extraction unit, and the plurality of process sensors are distributed in various areas of the chip;

[0031] Each process sensor is used to collect process parameters of a corresponding area and send the process parameters to the extreme value process extraction unit;

[0032] The extreme value process extraction unit is used to extract the minimum process parameter from a plurality of process parameters and send the minimum process parameter to the dynamic compensation module.

[0033] In a possible implementation, the dynamic compensation module further includes a feedback protection unit;

[0034] The feedback protection unit is used to set the current feedback voltage as the supply voltage to disable the supply voltage regulation function when a failure occurs in the adaptive control circuit or the chip system.

[0035] According to a second aspect of the present application, a method for adaptively controlling chip power consumption is provided, which is used in the adaptive control circuit for chip power consumption as described above, and the method includes:

[0036] Collect the temperature parameters of each area in the chip to obtain the minimum temperature parameter;

[0037] Collecting process parameters of each area in the chip to obtain minimum process parameters;

[0038] Generate a simulation formula for the minimum supply voltage surface of the chip, input the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and use the calculation result of the simulation formula as the target resistance value, wherein the minimum supply voltage surface is simulated according to the temperature extreme value and the process extreme value of the chip;

[0039] Adjusting the resistance value of the variable resistor to the target resistance value, and adjusting the current feedback voltage to the target feedback voltage according to the adjusted variable resistor;

[0040] The power supply voltage of the chip is adjusted according to the current feedback voltage.

[0041] According to a third aspect of the present application, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, and the at least one instruction is loaded and executed by a processor to implement the adaptive control method of chip power consumption as described above.

[0042] According to a fourth aspect of the present application, a chip is provided, wherein the chip includes the adaptive control circuit for the power consumption of the chip.

[0043] The beneficial effects of the technical solution provided by this application include at least:

[0044] A simulation formula is generated through the minimum power supply voltage surface of the chip, and then the minimum temperature parameters, minimum process parameters, target feedback voltage and preset calibration parameters are input into the simulation formula. The resistance value of the variable resistor is adjusted according to the calculation result of the simulation formula to make the current feedback voltage reach the target feedback voltage. Finally, the power supply voltage of the chip is adjusted according to the current feedback voltage. In this way, the power consumption of the chip can be adjusted based on the two factors of temperature and process. While reducing the power consumption in extreme scenarios, the difference in power consumption of different temperatures and different chips is effectively reduced, greatly improving the consistency of the system, allowing manufacturers to reduce the yield loss caused by screening out high-power chips, thereby increasing profit margins. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 This is a schematic diagram of the temperature rise curve of the static power consumption of a chip with a typical process;

[0047] Figure 2 It is a schematic diagram of the minimum supply voltage curve required to maintain normal operation of the chip;

[0048] Figure 3 This is a schematic diagram of the table of power consumption values ​​of chips with different processes at different temperatures under the same supply voltage (0.8v);

[0049] Figure 4 It is a schematic diagram of a table showing the minimum supply voltage required for chips of different processes at different temperatures under the same operating frequency;

[0050] Figure 5 It is a schematic diagram of the table of chip power consumption under various conditions after adjusting the supply voltage;

[0051] Figure 6 is a schematic diagram of an adaptive control circuit for chip power consumption provided by an embodiment of the present application;

[0052] Figure 7 It is a schematic diagram of the concave surface of the minimum supply voltage in three-dimensional space;

[0053] Figure 8 is a schematic diagram of a supply voltage plane simulated using a three-dimensional plane approximation method;

[0054] Fig. 9It is a schematic diagram of a table that calculates the power supply voltage required for each temperature and process based on an approximate plane;

[0055] Fig.10 is a schematic diagram of a table showing the error between the approximately flat supply voltage and the original minimum supply voltage;

[0056] Fig.11 is a schematic diagram of an adaptive control circuit for chip power consumption provided by an embodiment of the present application;

[0057] Fig.12 It is a schematic diagram of the power consumption of the system in various cases after the supply voltage is adjusted;

[0058] Fig.13 This is a flow chart of a method for adaptively controlling chip power consumption provided by an embodiment of the present application. DETAILED DESCRIPTION

[0059] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0060] The power consumption of a chip mainly consists of two parts: static power consumption Ps and dynamic power consumption Pd. Static power consumption Ps refers to the basic power consumption of the chip when it is not performing any switching activities or working state; dynamic power consumption Pd refers to the power consumption consumed by the internal transistor flipping when the chip is working. Although the dynamic power consumption Pd is less affected by the operating temperature, the static power consumption Ps shows an exponential increase trend with the operating temperature. In high temperature scenarios of 125° or even 150°, the static power consumption Ps will even be much higher than the dynamic power consumption Pd. Figure 1 This is the temperature rise curve of the static power consumption of a typical process (TT) chip produced based on the 22nm process.

[0061] At the same time, with the change of temperature, when the operating frequency remains unchanged, the minimum supply voltage Vmin required to maintain the normal operation of the chip shows a U-shaped change law. Figure 2 As shown, the minimum supply voltage Vmin at low temperature and high temperature is relatively high, the minimum supply voltage Vmin at an intermediate temperature is low, and the minimum supply voltage Vmin at low temperature is higher than the minimum supply voltage Vmin at high temperature.

[0062] The static power consumption Ps of the chip shows greater differences with the deviation of the process. Usually, the differences in the production process of wafers will cause different wafers to show speed differences. In the production process, TT is often used to characterize the target value of process control. TT-N characterizes that the wafer is slower than TT, and the larger the N, the slower the speed is compared with TT; TT+N characterizes that the wafer is faster than TT, and the larger the N, the faster the speed is compared with TT. When the wafer is biased towards TT-N, the operating frequency of the chip is lower. In order to achieve the same speed as TT, the power supply voltage of the chip is required to be higher. The static power consumption Ps at the same voltage is smaller than the static power consumption Ps of TT. When the wafer is biased towards TT+N, the operating frequency of the chip is higher. In order to achieve the same speed as TT, the power supply voltage of the chip is required to be lower. The static power consumption Ps at the same voltage is greater than the static power consumption Ps of TT. Figure 3 The table shows the power consumption values ​​of chips with different processes at different temperatures under the same supply voltage (0.8V). The unit of the data in the table is (mW).

[0063] from Figure 3 As can be seen from the table, with the change of temperature and the deviation of the process, the power consumption of the chip shows a huge difference. The difference between the power consumption in the maximum scenario and the power consumption in the minimum scenario is more than 4.6 times. After the chip is installed in the actual system, the huge deviation in power consumption poses a huge challenge to the consistency of the product. At the same time, in order to meet the needs of the chip with the largest deviation under extreme temperature, the power supply capacity of the system power supply needs to be improved in the actual system, which will also cause a huge increase in the system power supply and heat dissipation costs.

[0064] In order to alleviate the above problems, chip manufacturers can choose to screen out chips with faster processes during the production process to improve the consistency of chip power consumption, but doing so will reduce the chip yield rate, increase the cost of good products, and reduce profit margins.

[0065] from Figure 4 From the table, we found that as the temperature changes, the minimum supply voltage Vmin shows a trend of first decreasing and then increasing; under the same process, the minimum supply voltage Vmin at low temperature is higher than the minimum supply voltage Vmin at high temperature; as the process changes from slow to fast, the minimum supply voltage Vmin shows a downward trend. Based on the above characteristics, the power consumption of the chip can be optimized by dynamically adjusting the chip's supply voltage according to the chip's operating temperature and the chip's process characteristics.

[0066] We also found that the dynamic power consumption Pd of the chip is in a square relationship with the supply voltage. That is, if the dynamic power consumption of the chip is 800mW under a supply voltage of 0.8v, when the supply voltage is reduced to 0.6v, the dynamic power consumption Pd will be reduced to 800×(0.6 / 0.8) 2 = 450mW. Under the 22nm process, the static power consumption Ps and the power supply voltage show a 2.5-power relationship. That is, if the static power consumption of the chip is 1000mW under 0.8v power supply, when the power supply voltage is reduced to 0.6v, the dynamic power consumption Pd will be reduced to 1000×(0.6 / 0.8) 2.5 = 487mW.

[0067] It can be seen that the power consumption of the chip can be effectively reduced by adjusting the supply voltage. According to the minimum supply voltage Vmin of the chip under different operating temperatures and processes, after adjusting the supply voltage, the power consumption under various conditions is as follows: Figure 5 As shown in the table. Figure 5 It can be seen that under the same conditions, the power consumption of the chip after adjusting the power supply voltage is less than that before the adjustment. Overall, the system will be more power-efficient after adjusting the power supply voltage. For example, the maximum power consumption of the chip has dropped from 3709mW to 1341mW, which greatly reduces the demand for system power supply and reduces costs; and the ratio of maximum and minimum power consumption has dropped from 4.6 to 1.7, which greatly improves the consistency of the system; in addition, chip manufacturers can reduce the yield loss caused by screening out high-power chips, thereby increasing profit margins.

[0068] like Figure 6 As shown, it shows a structural block diagram of an adaptive control circuit for chip power consumption provided by an embodiment of the present application, and the adaptive control circuit for chip power consumption can be applied to a chip. The adaptive control circuit for chip power consumption may include: a temperature acquisition module 610, a process acquisition module 620, a dynamic compensation module 630, a voltage feedback module 640 and a power module 650. Among them, the temperature acquisition module 610 is respectively connected to the dynamic compensation module 630 and the voltage feedback module 640, the process acquisition module 620 is respectively connected to the dynamic compensation module 630 and the voltage feedback module 640, the voltage feedback module 640 is connected to the power module 650, and the power module 650 is connected to the on-chip power supply network of the chip to provide a power supply voltage V for the chip. IN Taking the 22nm process as an example, the supply voltage V IN The typical value is 0.8v.

[0069] The temperature acquisition module 610 is used to acquire temperature parameters of various regions in the chip and send the minimum temperature parameter to the dynamic compensation module 630. Specifically, the temperature acquisition module 610 can acquire temperature parameters in multiple regions in the chip, extract the minimum temperature parameter from these temperature parameters, and send the minimum temperature parameter to the dynamic compensation module 630.

[0070] The process acquisition module 620 is used to collect process parameters of various regions in the chip and send the minimum process parameter to the dynamic compensation module 630. Specifically, the process acquisition module 620 can collect process parameters in multiple regions in the chip, extract the minimum process parameter from these process parameters, and send the minimum process parameter to the dynamic compensation module 630.

[0071] The dynamic compensation module 630 is used to generate a simulation formula for the minimum power supply voltage surface of the chip, input the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and send the calculation result of the simulation formula as the target resistance value to the voltage feedback module 640. The minimum power supply voltage surface is simulated based on the temperature extremes and process extremes of the chip.

[0072] Within the working range of the temperature extreme value (-40° to 150°) and the production deviation of the process extreme value (±3σ), the minimum supply voltage Vmin of the chip presents the characteristics of a concave surface in three-dimensional space. Based on this characteristic, we can use three-dimensional plane approximation method, lookup table method, three-dimensional surface approximation method, etc. to simulate the minimum supply voltage surface and obtain a simulation formula. Then, the simulation formula can be used to calculate the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter to obtain the target resistance value, which is the target value of the variable resistor in the voltage feedback module 640.

[0073] The voltage feedback module 640 is used to adjust the resistance of the variable resistor to the target resistance, adjust the current feedback voltage to the target feedback voltage according to the adjusted variable resistance, and send the current feedback voltage to the power supply module 650. The purpose of adjusting the resistance of the variable resistor to the target resistance is to generate a target feedback voltage so as to adjust the supply voltage through the target feedback voltage.

[0074] The power module 650 is used to adjust the supply voltage of the chip according to the current feedback voltage. In this embodiment, the power module 650 is located inside the chip for illustration. In actual implementation, the power module 650 can also be located outside the chip.

[0075] To summarize, the adaptive control circuit of chip power consumption provided in the embodiment of the present application generates a simulation formula through the minimum power supply voltage surface of the chip, and then inputs the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and adjusts the resistance value of the variable resistor according to the calculation result of the simulation formula so that the current feedback voltage reaches the target feedback voltage. Finally, the power supply voltage of the chip is adjusted according to the current feedback voltage. In this way, the power consumption of the chip can be adjusted according to the two factors of temperature and process. While reducing the power consumption in extreme scenarios, the difference in power consumption of different temperatures and different chips is effectively reduced, which greatly improves the consistency of the system, enabling manufacturers to reduce the yield loss caused by screening out high-power chips, thereby improving profit margins.

[0076] The structures of the voltage feedback module 640 , the temperature acquisition module 610 , the process acquisition module 620 and the dynamic compensation module 630 are described below respectively.

[0077] (1) Voltage feedback module 640

[0078] like Figure 6 As shown, the voltage feedback module 640 includes an amplifier, a fixed resistor R1 and a variable resistor R2. The in-phase input terminal of the amplifier is respectively connected to the output terminal of the power module 650 and the input terminal of the on-chip power supply network, the inverting input terminal of the amplifier is respectively connected to one end of the fixed resistor R1 and the variable resistor R2, the other end of the fixed resistor R1 is grounded, the other end of the variable resistor R2 is connected to the output terminal of the amplifier and then connected to the input terminal of the power module 650, and the adjustment end of the variable resistor R2 is connected to the dynamic compensation module 630.

[0079] The input of the power module 650 is the feedback voltage V FB , the output is the supply voltage V IN , feedback voltage V FB The power supply feedback given by the chip. Feedback voltage V FB and supply voltage V IN Satisfies the following relationship: V FB = V IN × (1+R2 / R1), R2 is a variable resistor structure with a value range of [0~R1]. FB When setting the target value, the supply voltage V can be calculated according to the formula IN Assuming the target feedback voltage V FB is 0.8v, then the equation can be transformed to get V IN = 0.8×R1 / (R1 + R2), by adjusting R2, the supply voltage V can be adjusted within the range of [0.8v~0.4v] IN .

[0080] That is, in the voltage feedback module 640, an amplifier is used to generate the feedback voltage V FB , so that the power module 650 supplies the supply voltage V IN Adjust to maintain the minimum voltage under the current working condition of the chip to save power consumption.

[0081] In order to adjust the feedback voltage to the target feedback voltage V FB , the target resistance value of the variable resistor R2 needs to be calculated by the dynamic compensation module 630, as described below for details.

[0082] (2) Temperature acquisition module 610

[0083] The temperature acquisition module 610 includes a plurality of temperature sensors and an extreme temperature extraction unit. The plurality of separate temperature sensors are distributed in various regions of the chip so as to sense the temperature of various regions inside the chip.

[0084] Specifically, each temperature sensor is used to collect temperature parameters of the corresponding area and send the temperature parameters to the extreme temperature extraction unit; the extreme temperature extraction unit is used to extract the minimum temperature parameter Tmin from multiple temperature parameters and send the minimum temperature parameter Tmin to the dynamic compensation module 630.

[0085] Optionally, the adaptive control circuit further includes a fault detection module, and the extreme temperature extraction unit is further used to extract a maximum temperature parameter Tmax from the multiple temperature parameters, and send the maximum temperature parameter Tmax to the fault detection module; the fault detection module is used to detect whether to issue an over-temperature warning according to the maximum temperature parameter Tmax. If the fault detection module issues an over-temperature warning, the chip system can be shut down.

[0086] (3) Process acquisition module 620

[0087] The process acquisition module 620 includes a plurality of process sensors and an extreme value process extraction unit. The plurality of separate process sensors are distributed in various regions in the chip so as to sense the processes in various regions inside the chip.

[0088] Specifically, each process sensor is used to collect process parameters of the corresponding area and send the process parameters to the extreme value process extraction unit; the extreme value process extraction unit is used to extract the minimum process parameter Pslow from multiple process parameters and send the minimum process parameter Pslow to the dynamic compensation module 630. The minimum process parameter Pslow represents the slowest value of the process deviation of the internal circuit of the chip.

[0089] (4) Dynamic compensation module 630

[0090] The relationship between the minimum supply voltage Vmin of the chip (taking 22nm process as an example) and the temperature and process is as follows: Figure 7As shown. Within the working range of temperature extremes (-40° to 150°) and the production deviation of process extremes (±3σ), the minimum supply voltage Vmin of the chip presents the characteristics of a concave surface in three-dimensional space. Based on this characteristic, we can use three-dimensional plane approximation method, lookup table method, three-dimensional surface approximation method, etc. to simulate the minimum supply voltage surface.

[0091] In order to simplify the dynamic compensation algorithm and reduce the production test time, this embodiment uses a plane formed by four extreme points {(-40C, TT-3), (-40C, TT+3), (150C, TT-3), (150C, TT+3)} to approximate the minimum supply voltage surface. Figure 8 It can be seen that, except for the three extreme points, the voltage values ​​of all points on the plane are higher than the voltage values ​​in the concave surface. It is safe to adjust the supply voltage according to the plane parameters during the operation of the chip.

[0092] According to the four extreme points of the minimum supply voltage surface, the plane formula with the maximum Vmin is obtained: V = 0.74 - 0.000316 × (T + 40) - 0.0366667 × (P + 3), where V is the minimum supply voltage, T is the temperature parameter of the chip, and P is the process parameter of the chip (P is a linear parameter with the unit of σ).

[0093] Calculate the power supply voltage V required for each temperature and process based on the approximate plane IN like Fig. 9 As shown, the approximately flat supply voltage V IN The error between the original minimum supply voltage Vmin is as follows Fig.10 It can be seen that the results of using the approximate plane are not lower than the original minimum supply voltage Vmin, so the approximate plane can maintain the normal operation of the chip.

[0094] In this embodiment, the three-dimensional plane approximation method is taken as an example. In actual use, the formula needs to be calibrated. The calibration parameters come from the one-time programmer (OTP) in the chip. The obtained simulation formula is:

[0095] V IN=Vmin_ini – Tadj – Padj = Vmin_ini – Tgain× (Tmin + Toffset) –Pgain×(Pslow + Poffset), where Vmin_ini is the initial minimum voltage parameter, Tadj is the temperature adjustment parameter, Padj is the process adjustment parameter, Tgain is the temperature gain parameter, Toffset is the temperature offset parameter, Pslow is the process gain parameter, and Poffset is the process offset parameter.

[0096] That is, in the simulation formula, the supply voltage is equal to the initial minimum voltage parameter minus the temperature adjustment parameter and the process adjustment parameter, the temperature adjustment parameter is equal to the sum of the minimum temperature parameter and the temperature offset parameter multiplied by the temperature gain parameter, and the process adjustment parameter is equal to the sum of the minimum process parameter and the process offset parameter multiplied by the process gain parameter.

[0097] Since the resistance of the variable resistor R2 needs to be adjusted in the end, combined with the voltage feedback formula V IN =V FB × R1 / (R1+R2), we get R2 = V FB ×R1 / (Vmin_ini - Tadj - Padj) - R1.

[0098] In order to avoid over-limit during the adjustment process, the protection unit needs to limit the adjustment of the corresponding parameters. Fig.11 As shown, the dynamic compensation module 630 includes at least one of a temperature protection unit, a process protection unit, and a resistance protection unit.

[0099] When the dynamic compensation module 630 includes a temperature protection unit, the temperature protection unit is used to obtain a maximum temperature adjustment parameter and a minimum temperature adjustment parameter in the calibration parameters, and limit the temperature adjustment parameter to between the maximum temperature adjustment parameter and the minimum temperature adjustment parameter.

[0100] Specifically, when Tgain×(Tmin + Toffset) < Tdown, Tadj = Tdown; when Tgain×(Tmin + Toffset) > Tup, Tadj = Tup; in other cases, Tadj = Tgain×(Tmin +Toffset). Tdown and Tup are parameters programmed in OTP.

[0101] When the dynamic compensation module 630 includes a process protection unit, the process protection unit is used to obtain a maximum process adjustment parameter and a minimum process adjustment parameter in the calibration parameters, and limit the process adjustment parameter to between the maximum process adjustment parameter and the minimum process adjustment parameter.

[0102] Specifically, when Pgain×(Pslow + Poffset) < Pdown, Padj = Pdown; when Pgain×(Pslow + Poffset) > Pup, Padj = Pup; in other cases, Padj = Pgain×(Pslow +Poffset). Pdown and Pup are parameters programmed in OTP.

[0103] When the dynamic compensation module 630 includes a resistance protection unit, the resistance protection unit is used to obtain the maximum resistance value and the minimum resistance value in the calibration parameters, and limit the target resistance value between the maximum resistance value and the minimum resistance value.

[0104] Specifically, when R2 < R2down, R2adj = R2down; when R2 > R2up, R2adj = R2up; in other cases, R2adj = R2. R2down and R2up are parameters programmed in OTP.

[0105] Optionally, the dynamic compensation module 630 also includes a temperature validity check unit, which can detect the validity and legality of the temperature parameters measured by the temperature sensor. When the chip is working normally, its temperature range will not exceed the rated range. For example, the operating temperature of a chip that meets Automotive Grade-1 will not exceed the range of -40C to 150C. When it is detected that the temperature value exceeds the normal range, an output error signal (Error) is given to the temperature protection unit to turn off the temperature regulation function.

[0106] Specifically, the temperature validity check unit is used to detect whether the minimum temperature parameter is within the normal operating temperature range; if the minimum temperature parameter is not within the normal operating temperature range, an error signal is generated and sent to the temperature protection unit; the temperature protection unit is used to adjust the temperature gain parameter according to the error signal to turn off the temperature adjustment function. For example, when the temperature gain parameter Tgain is changed to 0, the temperature adjustment parameter Tadj is 0, which is equivalent to turning off the temperature adjustment function.

[0107] Optionally, the dynamic compensation module 630 also includes a process validity check unit, which compares the detection value of the process sensor with the initial process parameter Pinitial burned in the OTP. The chip will age during use, causing the process to gradually slow down, and a margin of σ is generally retained during the chip design process, so the process parameters detected when the chip is running will not be faster than one σ. If the process parameters detected during operation violate the above assumptions, an error signal (Error) will be generated and given to the process protection unit to turn off the process dynamic adjustment function. For example, when the process gain parameter Pgain is changed to 0, the process adjustment parameter Padj is 0, which is equivalent to turning off the process adjustment function. Among them, the initial process parameter Pinitial is the detection value of the process sensor during factory testing.

[0108] Specifically, the process validity check unit is used to detect whether the difference between the minimum process parameter and the initial process parameter exceeds a threshold value; if the difference between the minimum process parameter and the initial process parameter exceeds the threshold value, an error signal is generated and sent to the process protection unit; the process protection unit is used to adjust the process gain parameter according to the error signal to turn off the process adjustment function.

[0109] Optionally, the dynamic compensation module 630 further includes a feedback protection unit; the feedback protection unit is used to set the current feedback voltage to the supply voltage to disable the supply voltage regulation function when a system failure occurs in the adaptive control circuit or chip.

[0110] like Fig.11 As shown, the in-phase input terminal of the amplifier is connected to the first input terminal of the multiplexer Mux, the output terminal of the amplifier is connected to the second input terminal of the multiplexer Mux, the feedback protection unit is connected to the third input terminal of the multiplexer Mux, and the output terminal of the multiplexer Mux is used to output the feedback voltage V FB When the feedback protection unit does not detect a fault, the multiplexer Mux selects the output result of the output amplifier, that is, the target feedback voltage V FB ; When the feedback protection unit detects a fault, the multiplexer Mux selects the output supply voltage V IN , that is, setting the current feedback voltage to the supply voltage to turn off the supply voltage regulation function.

[0111] According to the three-dimensional plane approximation method, the supply voltage V IN After adjustment, the system power consumption in each case is as follows Fig.12 Compared with the ideal adjustment method, the power consumption in the extreme case is basically not increased, and the overall power consumption is slightly increased, but it is similar to the fixed supply voltage V IN Compared with the previous method, the power consumption optimization is still very obvious.

[0112] like Fig.13 As shown, it shows a flow chart of a method for adaptively controlling chip power consumption provided by an embodiment of the present application, and the method for adaptively controlling chip power consumption can be applied to a chip. The method for adaptively controlling chip power consumption can include:

[0113] Step 1301, collecting temperature parameters of various regions in the chip to obtain minimum temperature parameters.

[0114] The temperature acquisition module includes multiple temperature sensors and an extreme temperature extraction unit. Each temperature sensor acquires the temperature parameters of the corresponding area and sends the temperature parameters to the extreme temperature extraction unit; the extreme temperature extraction unit extracts the minimum temperature parameter Tmin from the multiple temperature parameters and sends the minimum temperature parameter Tmin to the dynamic compensation module.

[0115] Step 1302, collecting process parameters of each area in the chip to obtain minimum process parameters.

[0116] The process acquisition module includes multiple process sensors and an extreme process extraction unit. Each process sensor collects process parameters of the corresponding area and sends the process parameters to the extreme process extraction unit; the extreme process extraction unit extracts the minimum process parameter Pslow from multiple process parameters and sends the minimum process parameter Pslow to the dynamic compensation module.

[0117] Step 1303, generate a simulation formula for the minimum power supply voltage surface of the chip, input the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and use the calculation result of the simulation formula as the target resistance value. The minimum power supply voltage surface is simulated based on the temperature extremes and process extremes of the chip.

[0118] The simulation formula generated by the dynamic compensation module is: V IN =Vmin_ini – Tadj – Padj = Vmin_ini –Tgain× (Tmin + Toffset) – Pgain×(Pslow + Poffset), where Vmin_ini is the initial minimum voltage parameter, Tadj is the temperature adjustment parameter, Padj is the process adjustment parameter, Tgain is the temperature gain parameter, Toffset is the temperature offset parameter, Pslow is the process gain parameter, and Poffset is the process offset parameter.

[0119] Since the resistance of the variable resistor R2 needs to be adjusted in the end, combined with the voltage feedback formula V IN =V FB × R1 / (R1+R2), we get R2 = V FB×R1 / (Vmin_ini - Tadj - Padj) - R1.

[0120] Step 1304: adjust the resistance of the variable resistor to a target resistance, and adjust the current feedback voltage to a target feedback voltage according to the adjusted variable resistor.

[0121] The voltage feedback module adjusts the resistance value of the variable resistor to a target resistance value, so as to adjust the current feedback voltage to a target feedback voltage according to the adjusted variable resistor.

[0122] Step 1305: adjust the chip supply voltage according to the current feedback voltage.

[0123] Feedback voltage V FB and the supply voltage V IN Satisfies the following relationship: V FB = V IN × (1+R2 / R1), after obtaining the target feedback voltage and target resistance, the supply voltage V can be calculated. IN .

[0124] To summarize, the adaptive control method of chip power consumption provided in the embodiment of the present application generates a simulation formula through the minimum power supply voltage surface of the chip, and then inputs the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and adjusts the resistance value of the variable resistor according to the calculation result of the simulation formula so that the current feedback voltage reaches the target feedback voltage. Finally, the power supply voltage of the chip is adjusted according to the current feedback voltage. In this way, the power consumption of the chip can be adjusted according to the two factors of temperature and process. While reducing the power consumption in extreme scenarios, the difference in power consumption of different temperatures and different chips is effectively reduced, which greatly improves the consistency of the system, enabling manufacturers to reduce the yield loss caused by screening out high-power chips, thereby improving profit margins.

[0125] An embodiment of the present application provides a chip, wherein the chip includes an adaptive control circuit for power consumption of any of the above chips.

[0126] It should be noted that: the adaptive control circuit of chip power consumption provided in the above embodiment only uses the division of the above functional modules as an example when performing adaptive control of chip power consumption. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the adaptive control circuit of chip power consumption is divided into different functional modules to complete all or part of the functions described above. In addition, the adaptive control circuit of chip power consumption provided in the above embodiment and the adaptive control method embodiment of chip power consumption belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0127] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0128] The above description is not intended to limit the embodiments of the present application. Any adjustments, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the protection scope of the embodiments of the present application.

Claims

1. An adaptive control circuit for chip power consumption, characterized in that: The adaptive control circuit includes a temperature acquisition module, a process acquisition module, a dynamic compensation module, a voltage feedback module and a power supply module; The temperature acquisition module is used to acquire temperature parameters of various areas in the chip and send the minimum temperature parameter to the dynamic compensation module; The process acquisition module is used to collect process parameters of each area in the chip and send the minimum process parameter to the dynamic compensation module; The dynamic compensation module is used to generate a simulation formula for the minimum power supply voltage surface of the chip, input the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and send the calculation result of the simulation formula as the target resistance to the voltage feedback module, the minimum power supply voltage surface is simulated according to the temperature extreme value and the process extreme value of the chip, and the calibration parameters include an initial minimum voltage parameter, a temperature gain parameter, a temperature offset parameter, a process gain parameter and a process offset parameter; in the simulation formula, the power supply voltage is equal to the initial minimum voltage parameter minus the temperature adjustment parameter and the process adjustment parameter, the temperature adjustment parameter is equal to the sum of the minimum temperature parameter and the temperature offset parameter multiplied by the temperature gain parameter, and the process adjustment parameter is equal to the sum of the minimum process parameter and the process offset parameter multiplied by the process gain parameter; The voltage feedback module is used to adjust the resistance value of the variable resistor to the target resistance value, adjust the current feedback voltage to the target feedback voltage according to the adjusted variable resistor, and send the current feedback voltage to the power module; The power supply module is used to adjust the power supply voltage of the chip according to the current feedback voltage; The dynamic compensation module includes at least one of a temperature protection unit, a process protection unit and a resistance protection unit; the temperature protection unit is used to obtain a maximum temperature adjustment parameter and a minimum temperature adjustment parameter in the calibration parameters, and limit the temperature adjustment parameter between the maximum temperature adjustment parameter and the minimum temperature adjustment parameter; The process protection unit is used to obtain a maximum process adjustment parameter and a minimum process adjustment parameter in the calibration parameters, and limit the process adjustment parameter to between the maximum process adjustment parameter and the minimum process adjustment parameter; The resistance protection unit is used to obtain the maximum resistance value and the minimum resistance value in the calibration parameters, and limit the target resistance value between the maximum resistance value and the minimum resistance value; The dynamic compensation module further includes a temperature validity checking unit; the temperature validity checking unit is used to detect whether the minimum temperature parameter is within the normal operating temperature range; if the minimum temperature parameter is not within the normal operating temperature range, an error signal is generated and sent to the temperature protection unit; The temperature protection unit is used to adjust the temperature gain parameter according to the error signal to turn off the temperature adjustment function; The dynamic compensation module further includes a process validity checking unit; the process validity checking unit is used to detect whether the difference between the minimum process parameter and the initial process parameter exceeds a threshold value; if the difference between the minimum process parameter and the initial process parameter exceeds the threshold value, an error signal is generated and sent to the process protection unit; the process protection unit is used to adjust the process gain parameter according to the error signal to turn off the process adjustment function; The process acquisition module includes a plurality of process sensors and an extreme process extraction unit, wherein the plurality of process sensors are distributed in various regions of the chip; each process sensor is used to collect process parameters of a corresponding region and send the process parameters to the extreme process extraction unit; the extreme process extraction unit is used to extract a minimum process parameter from a plurality of process parameters and send the minimum process parameter to the dynamic compensation module; The dynamic compensation module also includes a feedback protection unit; the feedback protection unit is used to set the current feedback voltage to the supply voltage to disable the supply voltage regulation function when a failure occurs in the adaptive control circuit or the chip system.

2. The adaptive control circuit of chip power consumption according to claim 1, characterized in that: The adaptive control circuit further comprises a fault detection module, the temperature acquisition module comprises a plurality of temperature sensors and an extreme temperature extraction unit, and the plurality of temperature sensors are distributed in various areas of the chip; Each temperature sensor is used to collect temperature parameters of a corresponding area and send the temperature parameters to the extreme temperature extraction unit; The extreme temperature extraction unit is used to extract a maximum temperature parameter and a minimum temperature parameter from a plurality of temperature parameters, send the maximum temperature parameter to the fault detection module, and send the minimum temperature parameter to the dynamic compensation module; The fault detection module is used to detect whether to issue an over-temperature warning according to the maximum temperature parameter.

3. A method for adaptively controlling chip power consumption, characterized in that: Used in the adaptive control circuit of chip power consumption as claimed in claim 1 or 2, the method comprising: Collect the temperature parameters of each area in the chip to obtain the minimum temperature parameter; Collecting process parameters of each area in the chip to obtain minimum process parameters; Generate a simulation formula for the minimum power supply voltage surface of the chip, input the minimum temperature parameter, the minimum process parameter, the target feedback voltage and the preset calibration parameter into the simulation formula, and use the calculation result of the simulation formula as the target resistance value, the minimum power supply voltage surface is simulated according to the temperature extreme value and the process extreme value of the chip, and the calibration parameters include an initial minimum voltage parameter, a temperature gain parameter, a temperature offset parameter, a process gain parameter and a process offset parameter; in the simulation formula, the power supply voltage is equal to the initial minimum voltage parameter minus the temperature adjustment parameter and the process adjustment parameter, the temperature adjustment parameter is equal to the sum of the minimum temperature parameter and the temperature offset parameter multiplied by the temperature gain parameter, and the process adjustment parameter is equal to the sum of the minimum process parameter and the process offset parameter multiplied by the process gain parameter; Adjusting the resistance value of the variable resistor to the target resistance value, and adjusting the current feedback voltage to the target feedback voltage according to the adjusted variable resistor; The power supply voltage of the chip is adjusted according to the current feedback voltage.

4. A chip, characterized in that: The chip comprises: the adaptive control circuit of chip power consumption as claimed in claim 1 or 2.

Images