Processor skin temperature power management based on multiple external sensors

Abstract

A processing unit [110] manages temperature by correlating readings from a plurality of external temperature sensors [106, 107] to the skin temperature of the processing unit, wherein the correlation is based on characteristics of a computer chassis [223] to include the processing unit. The processing unit is mounted on a printed circuit board (PCB) [102] or other substrate to be placed in a computer chassis. Each of the plurality of temperature sensors is placed at a different location on the PCB to provide temperature readings from multiple locations on the PCB. A temperature controller [115] of the processing unit receives temperature readings from the plurality of sensors and correlates the temperature readings to the skin temperature of the processing unit based on a plurality of correlation values [118].

Description

Background Technology

[0001] To improve efficiency and reduce power consumption, computing devices sometimes employ processors with multiple heterogeneous processor cores, called Accelerated Processing Units (APUs). For example, some APUs include one or more Central Processing Unit (CPU) cores and one or more Graphics Processing Unit (GPU) cores. However, during operation, the multiple processor cores of an APU often generate a relatively large amount of heat, which can affect the reliability and lifespan of the APU. Therefore, computer systems including APUs typically employ temperature management systems that use components such as heat pipes to transfer heat from the APU to corresponding heat sinks for heat dissipation. However, traditional temperature management systems are relatively inflexible and have varying effectiveness in different computing environments. Attached Figure Description

[0002] This disclosure will be better understood by referring to the accompanying drawings, and many of its features and advantages will become apparent to those skilled in the art. The same reference numerals are used in different drawings to indicate similar or identical items.

[0003] Figure 1 This is a block diagram of a processing system with a temperature management system according to some embodiments, the temperature management system correlating temperature readings from multiple temperature sensors to manage the temperature of an accelerated processing unit (APU).

[0004] Figure 2 It is based on the impact of some implementation plans. Figure 1 A diagram of a computer chassis showing the temperature readings of a temperature sensor.

[0005] Figure 3 This is a chassis profile based on some implementation schemes for predicting the thermal impact of computer chassis. Figure 1 A block diagram of the system that generates correlation values.

[0006] Figure 4 This is a flowchart of a method for managing the temperature of an APU by using multiple temperature sensors in a processing system, based on some implementation schemes. Detailed Implementation

[0007] Figures 1 to 4A technique for managing the temperature of one or more processing units (such as an Accelerated Processing Unit (APU) or other heterogeneous processing units) is illustrated by correlating readings from multiple external temperature sensors (i.e., outside the processing unit) with the skin temperature of the processing unit, wherein the correlation is based on the characteristics of the computer chassis to which the processing unit is to be included. For example, in some embodiments, the processing unit is mounted on a printed circuit board (PCB) or other substrate to be placed within a computer chassis (e.g., a laptop computer chassis). Each of the multiple temperature sensors is placed at a different location on the PCB to provide temperature readings from multiple locations on the PCB. A temperature controller for the processing unit receives the temperature readings from the multiple sensors and correlates the temperature readings with an estimated or otherwise predicted skin temperature of the processing unit based on multiple correlation values ​​representing a measured or predicted correspondence between the temperature at the sensor and the skin temperature of the processing unit, wherein the correlation values ​​are based on the characteristics of the computer chassis. In response to a temperature measurement indicating that the skin temperature may exceed a specified thermal threshold, the temperature controller takes remedial measures, such as reducing the power supplied to the processing unit.

[0008] As described in this article, managing processing unit temperature based on correlated temperature readings from multiple external sensors supports more flexible and accurate temperature management. Specifically, temperature increases at different locations within the computing system will correspondingly increase the skin temperature of the processing unit. However, in many cases, it is difficult or costly to detect such increases using temperature sensors mounted on the processing unit's own skin. Furthermore, computer systems can generate relatively large temperature differences at different locations, making it difficult for a single temperature sensor to provide an inaccurate indication of the processing unit's skin temperature. Therefore, by employing multiple temperature sensors at different locations within the computer system and by correlating measurements from different temperature sensors with the predicted skin temperature of the processing unit, the computer system can effectively manage processing unit temperature under various conditions and system configurations.

[0009] In some implementations, the correlation value used by the processing unit is based on a virtual chassis value that reflects the predictive influence of the computer chassis on the correlation between temperature sensor readings and the processing unit's skin temperature. For illustration, in many cases, the PCB and processing unit are designed to be housed within a computer chassis (such as a laptop computer chassis) to protect the PCB and associated components, connect the processing unit to other components of the processing system, etc. The computer chassis influences factors such as airflow and cooling patterns, which in turn affect the correlation between a given temperature sensor reading on the PCB and the processing unit's skin temperature. By incorporating these influences into the correlation value, the processing unit improves the accuracy of skin temperature measurements, thereby enhancing the effectiveness of temperature management.

[0010] In some implementations, the predicted impact of a computer chassis is identified based on testing of the PCB and temperature sensors, with the PCB housed within the chassis (or similar enclosure). However, in some cases, performing such testing is prohibitively expensive. Therefore, in some implementations, correlation values ​​are set based on a virtual model of the chassis that predicts the impact of the computer chassis on the correlation. This improves the accuracy of the correlation values ​​while reducing the cost of developing them.

[0011] Go to Figure 1 A block diagram of a processing system 100 that implements temperature control based on measurements from multiple temperature sensors is depicted according to some embodiments. The processing system 100 includes multiple components that together support the execution of computer instructions. Thus, in various embodiments, the processing system 100 is embedded in an electronic device, such as a desktop computer, laptop computer, server, game console, smartphone, tablet computer, automobile, or other vehicle, and executes a set of instructions (e.g., a computer program) to perform tasks on behalf of the electronic device.

[0012] To execute instructions, the processing system 100 includes an APU 110 with multiple processor cores, each processor core executing instructions or operations assigned to it. In some cases, the instructions are general-purpose instructions for performing common tasks, such as executing an operating system, interfaced with one or more input / output devices (not shown), etc. In other cases, these operations are specialized operations, such as graphics operations (e.g., drawing objects for a display), vector processing operations, machine learning or artificial intelligence operations, etc. To efficiently execute different types of instructions and operations, the APU 110 includes different types of processor cores. For example, in Figure 1 In one embodiment, the APU 110 includes a central processing unit (CPU) core 112 for executing general-purpose instructions and an integrated graphics processing unit (iGPU) core 114 for performing graphics operations. In other embodiments, the APU 110 includes additional CPU cores, additional GPU cores, or any combination thereof. As an alternative to or addition to one or more CPU cores or one or more GPU cores, in some embodiments, the APU 110 includes Figure 1 Other types of processor cores, not shown, such as one or more vector processing cores, digital signal processing (DSP) cores, artificial intelligence (AI) cores, inference engine cores, etc., each are used to perform one or more corresponding types of operations. Similarly, unless otherwise explicitly stated, the reference to CPU core 112 or iGPU core 114 is similarly applicable to other types of processor cores.

[0013] To further support instruction execution, processing system 100 includes a discrete coprocessor, such as a discrete GPU (dGPU) 116, which includes circuitry for performing graphics and vector processing operations based on commands received from APU 110, at least in some operating modes of processing system 100. For example, in some embodiments, processing system 100 can be placed in different modes based on one or more factors, such as the power supply to processing system 100 (e.g., battery power versus grid power), the type of program executed at processing system 100, user-programmed quality settings of processing system 100, etc. For example, in some embodiments, processing system 100 uses iGPU core 114 for graphics operations when processing system 100 is powered by battery power, and uses dGPU 116 for graphics operations when processing system 100 is powered by grid power.

[0014] In the depicted implementation, APU 110 is powered by power supply 104. In some implementations, power supply 109 draws power from a wired power source (i.e., a power source powered by electricity supplied from the power grid), a battery, or a combination thereof. In some implementations, power supply 109 provides controllable power to APU 110. That is, the power supplied by power supply 109 can be controlled based on control signaling provided to the power supply. As further described below, adjustments to power supply 109 help control the temperature of APU 110 and dGPU 116, thereby protecting the reliability and lifespan of these units.

[0015] The APU 110 and dGPU 116 are mounted on a printed circuit board (PCB) 102. In some implementations, PCB 102 is a computer motherboard or similar PCB on which the APU 110 and dGPU 116 are mounted. Figure 1 Additional components of the processing system 100, not specifically shown, include one or more memory modules, network interface modules, etc. The PCB 102 also includes connectors, buses, conductive traces, etc., for connecting components of the processing system 100, including connecting the APU 110 to one or more other components.

[0016] To assist in managing the various components, the processing system 100 includes an electronic controller (EC) 105. EC 105 is an integrated circuit controller device that manages various aspects of the processing system 100, such as temperature measurement and control, as further described below. In some embodiments, EC 105 performs other operations, such as implementing commands provided by the Basic Input / Output System (BIOS) firmware executed at APU 110. For example, in some embodiments, in response to a reset or other conditions, APU 110 automatically executes a stored set of BIOS firmware (…). Figure 1(Not shown in the image). During execution, the BIOS firmware sends commands to the EC 105 to initialize the storage system (e.g., hard disk drive or flash drive), initialize the network interface, initialize communication with other processing systems, etc. The EC 105 receives and interprets the commands and implements them by providing corresponding control signaling to different components of the processing system 100. The EC 105 thus provides an interface between the components of the processing system 100 and the APU 110, thereby allowing the APU 110 to be used with various processing systems with different configurations and components.

[0017] As described above, one aspect of the processing system 100 managed by EC 105 is the skin temperature of APU 110 and dGPU 116. For illustration, during operation of the processing system 100, various system components, including APU 110 and dGPU 116, generate heat. If the heat generated at or near APU 110 or dGPU 116 becomes excessively high for an extended period, the reliability or lifespan of APU 110 or dGPU 116 is adversely affected. Therefore, the processing system 100 employs a temperature control subsystem to manage the temperature of both APU 110 and dGPU 116, specifically maintaining the skin temperature of each processing unit below a corresponding threshold, above which the corresponding processing unit may be damaged or otherwise negatively affected. The temperature control subsystem includes EC 105, multiple temperature sensors 106, 107, and 108, and a temperature control module 115.

[0018] Each of temperature sensors 106 to 108 is a device that generates an electrical signal having an amplitude or other characteristic proportional to the temperature at or near the device. Examples of such devices include thermocouple sensors, resistance temperature detectors (RTDs), thermistor temperature sensors, integrated circuit (IC) temperature sensors, and combinations thereof. EC 105 receives electrical signals from temperature sensors 106 to 108 and generates a corresponding digital value for each temperature sensor based on the corresponding electrical signal. Therefore, EC 105 generates a digital value (referred to herein as a digital temperature value) representing the temperature at or near the corresponding sensor for each of temperature sensors 106 to 108. In some embodiments, including Figure 1 In the illustrated embodiment, temperature sensors 106 to 108 are placed at different locations on PCB 102. Therefore, the digital temperature values ​​generated by EC 105 represent the temperature at each location of the processing system 100.

[0019] It should be understood that the thermal characteristics at one location of the processing system 100 are based on heat sources, convection effects, etc., that have a general impact on the processing system 100 as a whole. Therefore, the thermal characteristics and temperature at one location of the processing system 100 are generally related to the thermal characteristics and temperature at another location of the processing system 100. The temperature control subsystem of the processing system 100 uses this property to determine the skin temperature of the APU 110 and the skin temperature of the dGPU 116. Specifically, the APU 110 includes a temperature control module 115 that stores sensor correlations 118, wherein a subset of the sensor correlations 118 indicates the correlation between the skin temperature of the APU 110 and the temperature at or near temperature sensors 106 to 108, and another subset of the sensor correlations 118 indicates the correlation between the skin temperature of the dGPU 116 and the temperature at or near temperature sensors 106 to 108. The temperature control module 115 receives digital temperature values ​​from EC 105 and applies the corresponding sensor correlation to the digital temperature values ​​to generate skin temperature values ​​for each of the APU 110 and the dGPU 116. For example, in some implementations, the temperature control module 115 generates the skin temperature value according to the following formula:

[0020] SK1 = m1x1 + m2x2 + ... + m n x n +C

[0021] Where SK1 is the skin temperature value, x Y It is the digital temperature value of sensor Y, m Y It is a digital temperature value x Y The sensor correlation value, and C is a constant based on the characteristics of PCB 102.

[0022] To determine the skin temperature of APU 110, the sensor correlation value applied to a given digital temperature value is a correlation coefficient indicating the correlation between the skin temperature of APU 110 and the temperature sensor that generates the corresponding digital value. For example, in some embodiments, m1 is a sensor correlation indicating the correlation between the temperature at or near temperature sensor 107 and the skin temperature of APU 110, and x1 is a digital temperature value indicating the temperature measured by temperature sensor 107. It should be understood that in some embodiments, the thermal characteristics of different locations of processing system 100 vary due to, for example, variations in surrounding materials and components, variations in airflow and convection, etc. Therefore, the correlation between the temperature measured by different temperature sensors and the skin temperature of APU 110 changes, and thus the value of the sensor correlation also changes. For example, in some embodiments, the value of sensor correlation m1 is different from the value of sensor correlation m2. Furthermore, it should be understood that in some embodiments, temperature control module 115 uses the above formula to identify the skin temperature of dGPU 116, but with a different set of correlation values ​​(i.e., different values ​​of m1, m2, etc.) and a different constant C than that used for APU 110. The identification of sensor correlation 118, constant C, and other values ​​adopted by temperature control module 115 will be described further below.

[0023] In some implementations, the SK1 value is relatively noisy for individual skin temperature measurements of a given processing unit. Therefore, in these implementations, the temperature control module 115 applies an α filter to filter out noise according to the following formula:

[0024] SK α =SK1α+SK α-1 (1-α)

[0025] Among them SK α This is a filtered skin temperature value, α is an α-filtered value identified during the characterization process described further below, and SK α-1 It is the skin temperature value after filtering.

[0026] The temperature control subsystem of processing system 100 uses filtered skin temperature values ​​to control the skin temperature of APU 110 and dGPU 116. For example, in some embodiments, temperature control module 115 periodically compares each of the filtered skin temperature values ​​for APU 110 and dGPU 116 with a temperature threshold 119. In response to a filtered skin temperature exceeding the temperature threshold 119, temperature control module 115 reduces the power supplied to APU 110 by power supply 104, thereby ensuring that the skin temperature of APU 110 and dGPU 116 remains within specified safety limits. For example, in some embodiments, temperature control module 115 reduces the power to a specified amount. In other embodiments, the temperature control module reduces the power based on an amount proportional to the amount by which the filtered skin temperature exceeds the temperature threshold 119. In some embodiments, temperature control module 115 uses different temperature thresholds for APU 110 and dGPU 116.

[0027] As described above, in some embodiments, the values ​​of sensor correlation 118 (e.g., m1, m2, ..., mn) are determined during the characterization process of processing system 100. n The values ​​of the sensors are: a constant value C and a filter value α. For example, in some embodiments, PCB 102 (or a PCB of similar design and configuration) is placed in the test environment, where a temperature sensor is placed on the surface of the integrated circuit containing APU 110. During characterization, the temperature sensor provides an indication of the skin temperature of APU 110. Different test workloads are performed at APU 110, and for each executed workload, a set of temperature readings is acquired from each of temperature sensors 106 to 108. These temperature readings are correlated with the skin temperature measurement of APU 110 to determine the sensor correlation set. A similar process is used to determine the sensor correlation set of dGPU 116.

[0028] In some implementations, the sensor correlation of the APU 110 and dGPU 116 is further adjusted based on the characteristics of the computer chassis in which the PCB 102 is intended to be placed. For illustration, in many cases, the processing system 100 will be implemented within a computer chassis (such as a laptop computer chassis). According to some implementations... Figure 2 An exemplary computer chassis 220 is shown. In the depicted example, the computer chassis 220 includes a housing 223 containing multiple components. Therefore, in Figure 2In the exemplary arrangement, PCB 102 is placed in a slot near other computer components (e.g., component 222), such as disk drives, network interface cards, batteries or other power supply components, mechanical components for securing PCB 102 to housing 223, etc. In some cases, housing 223 and other components of computer chassis 220 affect the thermal behavior of processing system 100, particularly the correlation between temperature measurements at temperature sensors 106 to 108 and the skin temperatures of APU 110 and dGPU 116. Therefore, to improve the accuracy of skin temperature measurements performed by temperature control module 115, sensor correlation 118 is based on the characteristics of chassis 220.

[0029] For illustrative purposes, in some implementations, based on the operation of the PCB in a test bench environment, a given sensor correlation m has a nominal value, specified as m. 测试 However, test benches typically provide an environment very different from that of the computer chassis 220 and do not accurately simulate the thermal effects exerted by the computer chassis 220. Therefore, as further described herein, m is adjusted based on the expected effects of the computer chassis 220. 测试 This improves the accuracy of the skin temperature value generated by the temperature control module 115, thereby enhancing the effectiveness of the temperature control subsystem.

[0030] Figure 3 This is a block diagram of a system 300 for generating correlation values ​​for a processing system 100 based on a chassis profile predicting the thermal effects of a computer chassis 220, according to some embodiments. In some embodiments, system 300 is a computer system, such as a server or workstation, used during the characterization of PCB 102. System 300 executes a PCB configuration tool 330 to generate sensor correlations 118, as further described below. In some embodiments, PCB configuration tool 330 is a software tool that provides a graphical user interface (GUI) or other interface allowing users to adjust individual sensor correlation values ​​to test PCB 102 using different sensor correlation values ​​to determine the effectiveness of different values ​​for temperature control, thereby adjusting other test parameters such as temperature threshold 119.

[0031] To generate the sensor correlation 118 value, the PCB configuration tool 330 employs a chassis profile 332 and a sensor profile 334. In some embodiments, the sensor profile 334 stores the sensor correlation value data file based on testing of the PCB 102 in a test environment. For example, in some embodiments, the sensor profile 334 stores temperature measurement information as follows:

[0032]

[0033] Each measurement in the table above represents a different set of temperature measurements performed at PCB 102 under different workloads.

[0034] In some implementations, the PCB configuration tool 330 generates an initial sensor correlation value for each sensor based on the average of the correlations between the APU skin temperature measurements and the corresponding temperature measurements of the corresponding sensors, or other statistical combinations thereof. For example, the PCB configuration tool 330 generates an initial sensor correlation value for temperature sensor 106 based on the average of the correlations between temperatures T7 and T1 and the correlations between temperatures T8 and T4.

[0035] PCB configuration tool 330 then adjusts the initial sensor correlation values ​​based on chassis profile 332. In some embodiments, chassis profile 332 is a data file generated by testing the thermal characteristics of chassis 220 or another computer chassis. In other embodiments, chassis profile is based on a computer model of chassis 220. In other embodiments, chassis profile is based on testing or simulation of various aspects of chassis 220, such as one or more of the following: the shape of housing 223, components intended to be used with computer chassis 220 (e.g., the number of PCBs to be placed in slots in chassis 220), etc. Therefore, chassis profile 332 represents the predicted impact of computer chassis 220 on the sensor correlation values ​​of PCB 102. In some embodiments, chassis profile 332 stores these predicted impacts as adjustment factors for each of temperature sensors 106 to 108. An example is illustrated in the table below:

[0036] Sensor 106 Sensor 107 Sensor 108

[0037] A1 A2 A3

[0038] Where A1, A2, and A3 are adjustment factors corresponding to the initial correlation values ​​of temperature sensors 106, 107, and 108, respectively. To determine the sensor correlation value of a given temperature sensor, the PCB configuration tool 330 multiplies the initial correlation value of the sensor by the corresponding adjustment factor. Therefore, if m 106-I If the initial sensor correlation value of temperature sensor 106 is used, then PCB configuration tool 330 is based on the product A1*m 106-I The final correlation value is generated. The PCB configuration tool 330 stores the final correlation value at sensor correlation 118 for use during the operation of PCB 102.

[0039] It should be understood that, in some cases, the thermal effect of the computer chassis 220 varies depending on the location of the PCB 102. Therefore, the computer chassis 220 has different effects on the correlation between the skin temperature of the APU 110 and the temperature at different sensors among temperature sensors 106 to 108. Consequently, in some embodiments, the set of adjustment factors represented by the chassis profile 332 is different. For example, in some embodiments, adjustment factor A1 differs from adjustment factor A2.

[0040] In some implementations, a temperature sensor mounted on the surface of the dGPU 116 or on the skin is used to generate sensor correlation values ​​for the dGPU 116 in a manner similar to that described above with respect to the APU 110.

[0041] By employing chassis profile 332 to generate sensor correlation 118, PCB configuration tool 330 improves the accuracy of temperature measurement and temperature management during PCB 102 operation. Furthermore, in some embodiments, chassis profile 332 is generated based on computer modeling of computer chassis 220 or on thermal testing of a test PCB with a different configuration from PCB 102. This allows for the relatively inexpensive generation of chassis profile 332 without placing PCB 102 in computer chassis 220 for testing, thereby facilitating an efficient characterization process.

[0042] Figure 4 This is a flowchart of a method 400 for determining the skin temperature of a processing unit based on computer chassis information, according to some implementation schemes. (About...) Figure 1 An exemplary implementation of method 400 at APU 110 is described. At block 402, temperature control module 115 of APU 110 receives digital temperature values ​​from each of temperature sensors 106 to 108 from EC 105. At block 404, temperature control module 115 uses sensor correlation 118 to correlate the digital temperature values ​​with the skin temperature of APU 110 to generate a skin temperature value. As described above, sensor correlation 118 is based on the predicted thermal effects of computer chassis 220.

[0043] At box 406, temperature control module 115 determines whether the skin temperature value exceeds temperature threshold 119. If not, the method flow returns to box 402 and temperature control module 115 continues to monitor the skin temperature of APU 110. In response to the skin temperature value exceeding temperature threshold 119, the method flow moves to box 408 and temperature control module 115 indicates a temperature control event to EC 105. In response, the temperature control module reduces the power supplied to APU 110 by power supply 104, thereby maintaining the skin temperature of APU 110 within specified limits. The method flow returns to box 402.

[0044] In some embodiments, a method includes: receiving, at a first processing unit disposed in a computer chassis, a first temperature reading from a first temperature sensor external to the processing unit; adjusting the first temperature reading based on a first correlation value, the first correlation value being based on predicted temperature behavior of the computer chassis; and adjusting power supplied to the first processing unit based on the adjusted first temperature reading. In one aspect, the method includes: receiving, at the first processing unit, a second temperature reading from a second temperature sensor external to the first processing unit; adjusting the second temperature reading based on a second correlation value, the second correlation value being based on the predicted temperature behavior of the computer chassis; and adjusting power settings including adjusting the supplied power based on the adjusted second temperature reading. In another aspect, the second correlation value is different from the first correlation value.

[0045] In one aspect, the first correlation value is further based on an identified temperature correlation between the location of the first sensor and a location on the surface of the first processing unit. In another aspect, the location of the first sensor is a first location on a printed circuit board. In yet another aspect, the second correlation value is further based on an identified temperature correlation between a second location of the second sensor and a location on the surface of the processing unit, wherein the second location is a second location on the printed circuit board that is different from the first location. In yet another aspect, the method includes: adjusting the first temperature reading based on the second correlation value to identify a skin temperature associated with the second processing unit, the second correlation value being different from the first correlation value; and adjusting the power supplied to the first processing unit based on the identified skin temperature associated with the second processing unit.

[0046] In some implementations, a method includes: receiving at a first processing unit multiple temperature readings from multiple sensors external to the processing unit; identifying a skin temperature of the first processing unit by adjusting each of the multiple temperature readings based on a first plurality of correlation values, the first plurality of correlation values ​​being based on predicted temperature behavior of a computer chassis; and adjusting power supplied to the first processing unit based on the identified skin temperature of the first processing unit. In one aspect, the plurality of correlations includes a first correlation value and a second correlation value, the second correlation value being different from the first correlation value. In another aspect, each of the plurality of correlation values ​​is further based on an identified temperature correlation between the location of a sensor and a corresponding location on the surface of the processing unit.

[0047] In one aspect, the plurality of sensors are located at different locations on a printed circuit board. In another aspect, identifying the skin temperature includes filtering the skin temperature based on a previous skin temperature value. In yet another aspect, the skin temperature of the second processing unit is identified by adjusting each of the plurality of temperature readings based on a second plurality of correlation values ​​that are different from the first plurality of correlation values; and the power supplied to the first processing unit is adjusted based on the identified skin temperature of the second processing unit.

[0048] In some embodiments, a processing unit includes: a plurality of processor cores; a temperature control module configured to: receive a first temperature reading from a first temperature sensor external to the processing unit; adjust the first temperature reading based on a first correlation value, the first correlation value being based on predicted temperature behavior of a computer chassis; and initiate an adjustment of power supplied to the processing unit based on the adjusted first temperature reading. In one aspect, the temperature control module is configured to: receive a second temperature reading from a second temperature sensor external to the processing unit; adjust the second temperature reading based on a second correlation value, the second correlation value being based on the predicted temperature behavior of the computer chassis; and request power adjustment including requesting power adjustment based on the adjusted second temperature reading.

[0049] In one aspect, the second correlation value differs from the first correlation value. In another aspect, the first correlation value is also based on an identified temperature correlation between the location of the first sensor and a location on the surface of the processing unit. In yet another aspect, the first location is a first location on a printed circuit board. In yet another aspect, the second correlation value is also based on an identified temperature correlation between a second location of the first sensor and the location on the surface of the processing unit, wherein the second location is a second location on the printed circuit board that differs from the first location. In yet another aspect, the predicted temperature behavior is a simulated behavior of the computer chassis.

[0050] In some implementations, certain aspects of the techniques described above may be implemented by one or more processors of a processing system executing software. The software includes one or more sets of executable instructions stored or otherwise tangibly embodied on a non-transitory computer-readable storage medium. The software may include instructions and certain data that, when executed by one or more processors, manipulate the one or more processors to perform one or more aspects of the techniques described above. The non-transitory computer-readable storage medium may include, for example, disk or optical disk storage devices, solid-state storage devices such as flash memory, caches, random access memory (RAM), or one or more other non-volatile memory devices. The executable instructions stored on the non-transitory computer-readable storage medium may be in the form of source code, assembly language code, object code, or other instruction formats that are interpreted or otherwise executed by one or more processors.

[0051] It should be noted that not all activities or elements described in the general description above are required. A particular activity or apparatus may be unnecessary and may perform one or more other activities, or may include elements other than those described. Furthermore, the order in which the activities are listed is not necessarily the order in which they are performed. Additionally, the corresponding concepts have been described with reference to specific embodiments. However, those skilled in the art will understand that various modifications and changes can be made without departing from the scope of this disclosure as set forth in the appended claims. Therefore, this specification and drawings are to be regarded as illustrative rather than restrictive, and all such modifications are intended to be included within the scope of this disclosure.

[0052] The benefits, other advantages, and solutions to the problems have been described above with reference to specific embodiments. However, the benefits, advantages, and solutions to the problems described, as well as any features that may bring about or make more prominent any benefit, advantage, or solution to the problem, should not be construed as being key, essential, or necessary features of any or all claims. Furthermore, the specific embodiments disclosed above are merely illustrative, as the disclosed subject matter can be modified and practiced in different but equivalent ways that will be readily apparent to those skilled in the art who benefit from the teachings herein. No limitation is intended to be made on the details of the constructions or designs shown herein other than those described in the appended claims. Therefore, it is apparent that the specific embodiments disclosed above can be altered or modified, and all such changes are considered to be within the scope of the disclosed subject matter. Therefore, the protection sought herein is as set forth in the appended claims.

Claims

1. A method comprising: A first temperature reading is received at a first processing unit [110] located in a computer chassis [223] from a first temperature sensor [106] outside the first processing unit; The first temperature reading is adjusted based on a first correlation value [118] to generate an adjusted first temperature reading corresponding to the skin temperature of the first processing unit, the first correlation value being based on the predicted temperature behavior of the computer chassis. as well as The power supplied to the first processing unit is adjusted based on the adjusted first temperature reading.

2. The method of claim 1, further comprising: The first processing unit receives a second temperature reading from a second temperature sensor [107] located outside the first processing unit; The second temperature reading is adjusted based on a second correlation value to generate an adjusted second temperature reading corresponding to the second skin temperature of the first processing unit, the second correlation value being based on the predicted temperature behavior of the computer chassis; and The adjustment of power settings includes adjusting the supplied power based on a second temperature reading.

3. The method of claim 2, wherein the second correlation value is different from the first correlation value.

4. The method of claim 2 or claim 3, wherein the first correlation value is further based on an identified temperature correlation between the position of the first temperature sensor and the position at the surface of the first processing unit.

5. The method of claim 4, wherein the position of the first temperature sensor is a first position of the printed circuit board [102].

6. The method of claim 5, wherein the second correlation value is further based on an identified temperature correlation between a second position of the second temperature sensor and a position on the surface of the first processing unit, and wherein the second position is a second position on the printed circuit board that is different from the first position.

7. The method of claim 1, further comprising: The first temperature reading is adjusted based on a different correlation value than the first correlation value to identify the skin temperature associated with the second processing unit; as well as The power supplied to the first processing unit is adjusted based on the identified skin temperature associated with the second processing unit.

8. A processing unit [110], comprising: Multiple processor cores [112, 114]; Temperature control module [115], used for: Receive a first temperature reading from a first temperature sensor [106] outside the processing unit; The first temperature reading is adjusted based on a first correlation value [118] to generate a first adjusted temperature reading corresponding to the skin temperature of the processing unit, the first correlation value being based on the predicted temperature behavior of the computer chassis [223]. as well as Adjustments to the power supplied to the processing unit are initiated based on the first temperature reading.

9. The processing unit of claim 8, wherein the temperature control module is used to: Receive a second temperature reading from a second temperature sensor [107] outside the processing unit; The second temperature reading is adjusted based on a second correlation value to generate a second adjusted temperature reading corresponding to the second skin temperature of the processing unit, the second correlation value being based on the predicted temperature behavior of the computer chassis; and The request to adjust the power includes requesting an adjustment of the power based on a second temperature reading.

10. The processing unit of claim 9, wherein the second correlation value is different from the first correlation value.

11. The processing unit of claim 9 or claim 10, wherein the first correlation value is further based on an identified temperature correlation between the position of the first temperature sensor and the position at the surface of the processing unit.

12. The processing unit of claim 11, wherein the position of the first temperature sensor is a first position of the printed circuit board [102].

13. The processing unit of claim 12, wherein the second correlation value is further based on an identified temperature correlation between the position of the second temperature sensor and the position at the surface of the processing unit, and wherein the position of the second temperature sensor is a second position on the printed circuit board that is different from the first position.

14. The processing unit of claim 8, wherein the predicted temperature behavior is a simulated behavior of the computer chassis.

Images