TWI226530B - Method of reducing power consumption in processing unit - Google Patents

Abstract

A method of reducing consumption in processing unit. In the first, while a processing unit to start operating, sets a throttling level of the processing unit, that can set in any throttling level except a maximal utility rate of the processing unit. Then, detect the temperature of the processing unit under the said throttling level. Or detects the temperature and the ambient temperature of the processing unit, then calculate the difference of the temperature between the two. Finally, regulates out the throttling level of the processing unit fittingly, which according to the detected temperature of the processing unit, or calculated the difference of the temperature.

Description

1226530 V. Description of the invention (1) Technical field to which the invention belongs The present invention relates to a method for reducing energy consumption of a computer system, and in particular to a technology for reducing energy consumption of a processor in a computer system. In recent years, with the integration of integrated circuits (Integrated Circuit, IC) chip internal circuits (Integrati ο η) has been increasing, the central processing unit (CPU), The speed of integrated circuit (1C) chips such as Northbridge and graphics chips has also increased. In this way, the heating power inside the computer system is constantly rising. In order to prevent the integrated circuit chip in the computer brain system from overheating, the integrated circuit chip may temporarily or permanently fail, so in the computer system It becomes very important to provide adequate heat dissipation. Taking the central processing unit (Brai η) in the computer system as an example, the central processing unit operates at high speed. When the temperature of the central processing unit exceeds its normal operating temperature range, the central processing unit is extremely useful. Operation errors may occur or temporarily fail, which will cause the computer system to crash (Crash). In addition, it is very likely that the transistors inside the CPU chip will be damaged, which will cause the CPU to physically and permanently fail, and the CPU must be replaced to restore normal computer system operation. In order to effectively reduce the damage caused by the large amount of thermal energy φ generated by the integrated circuit chip during operation, in addition to installing a cooling fan inside the mainframe of the computer system, it can provide heat dissipation in the form of thermal convection.

1 1597t.wf.ptd Page 6 1226530 V. Description of the invention (2) Power supply of computer system, central processing unit, graphics processing unit (GPU), and even temperature rise products such as chipsets A device such as a heat sink is installed on the body circuit chip to directly contact the surface of the body circuit chip to provide a large heat dissipation area, so as to quickly remove the heat energy generated by the body circuit chip at high speed operation. Therefore, the temperature of the integrated circuit chip itself is reduced, which will make the operation of the computer system smoother. In computer systems, the thermal energy generated during the operation of integrated circuit chips can be said to be the factor that consumes the most energy during computer processing operations. Among them, the central processor is the most. The specifications of the CPU can usually be expressed in a few MHz of the Clock Frequency. This is to indicate how many basic actions the CPU performs in 1 second. Generally speaking, the higher the clock frequency, the higher the processing speed. Faster. In addition, 32-bit and 64-bit numbers are also used to indicate the specifications of the central processing unit. This refers to the amount of data that can be processed in one operation. The faster the clock pulse frequency, or the larger the number of bits, is not necessarily better, and different specifications of CPUs should be used depending on the purpose. However, even if an appropriate central processing unit is selected, during the period of use, there is a difference between the peak and low peak usage rates of the central processing unit. However, the existing technology can only fix the processor utilization ratio to a certain value. In the conventional technology, there is a central processor throttling function (CPU Throttling Func ο η) to achieve the purpose of reducing energy consumption. Please refer to FIG. 41 which shows the relationship between the processor usage rate and temperature of the conventional technology Schematic diagram, as shown in the general situation, the CPU throttling function is

1 1597t.wf.ptd Page 7 1226530 V. Description of the invention (3) In the disabled state, the CPU usage rate is set to 100%. When the system temperature is higher than a certain rated temperature, the central The processor throttling function is enabled to reduce the usage rate of the central processing unit to reduce the system temperature and protect the central processing unit from being damaged by overheating. Because this throttling function has only two levels, one is enabled and the other is disabled, and it cannot accurately achieve the purpose of reducing energy consumption. In short, because the conventional technology cannot dynamically change the duty cycle of the central processing unit and cannot meet the actual situation of using a computer system, it is impossible to accurately reduce energy consumption. SUMMARY OF THE INVENTION Therefore, the object of the present invention is to provide a method that can reduce the energy consumption of a processor in accordance with the actual use conditions. By reducing the energy consumption, thereby extending the use time of the notebook computer system in the battery mode and reducing the desktop The temperature of the computer in the idle state extends the service life of the internal integrated circuit chip. The present invention is based on the premise that the throttling degree of the processor is divided into a plurality of layers. First, when the processor starts to operate, the throttling degree of the processor is set in these layers in addition to the highest processor utilization rate. Any stratum outside the stratum. The next step is to detect the processor temperature at this level. Finally, the throttling degree of the processor is adjusted correspondingly according to the detected processor temperature. Among them, when the detected processor temperature falls within a certain predetermined value range, the throttling degree of the processor is fixed at a predetermined level corresponding to the predetermined value range. From another point of view, the present invention can also be first of all, when the processor starts to operate, the throttling degree of this processor is set in these levels except

1 1597t.wf.ptd Page 8 1226530 V. Description of the invention (4) Any level other than the level with the highest processor utilization. Next, detect the temperature of the processor and the temperature of the system environment where the processor is located (referred to as the system temperature). Finally, according to the detected temperature difference between the processor temperature and the system temperature, the throttling degree of the processor is adjusted accordingly. Wherein, when the temperature difference between the processor temperature and the system temperature falls within a predetermined value range, the throttle degree of the processor is fixed at a predetermined level corresponding to the predetermined value range. The invention adjusts the throttling degree according to the detected processor temperature, or changes the throttling degree of the processor according to the temperature difference between the detected processor temperature and the ambient temperature where the processor is located, so as to Multiple choices of throttling levels are used to effectively reduce unnecessary energy consumption of the processor without compromising actual use. ^ In order to make the above and other objects, features, and advantages of the present invention more comprehensible, a preferred embodiment is given below in conjunction with the accompanying drawings for detailed description as follows: Implementation: Please refer to FIG. 2 It shows a schematic flow chart according to the first embodiment of the present invention. This embodiment is applicable to a case where the throttling degree of the processor is divided into a plurality of levels. First, when the processor starts the arithmetic processing after the computer system is turned on, the throttling degree of the processor is set to the highest level ( Refers to any level other than the level with the highest processor utilization (eg, steps S 2 0, S 2 0 6). Next, the temperature at which the processor is operating under this level is detected (step S209). The last step is to adjust the throttling degree of the processor accordingly according to the detected processor temperature (eg step S 2 1 2

1 1597t.wf. Ptd page 9 1226530 V. Description of the invention (5)). Wherein, before step S 2 1 2, it is necessary to first set the corresponding relationship between the throttling degree hierarchy and the processor temperature required to adjust the throttling degree of the processor (not shown in the figure), and the adjustment method is: When the detected temperature of the processor is within a certain predetermined value range, the throttling degree of the processor is fixed at a predetermined level corresponding to the predetermined value range. Next, please refer to FIG. 3, which illustrates a schematic flow chart according to the second embodiment of the present invention. This embodiment is used in a situation where the throttling degree of the processor is divided into a plurality of levels. First, after the computer system is turned on, when the processor starts arithmetic processing, the throttling degree of the processor is set to the highest level. (Refers to the stratum with the highest processor utilization) (for example, steps S303, S306). Next, the temperature at which the processor is operating at this time is detected (step S309). The next step is to detect the temperature in the system environment where the processor is located at this time (step S 3 1 2). Finally, according to the detected temperature difference between the processor temperature and the system temperature, the throttling degree of the processor is adjusted correspondingly (step S31 5). Before step S3 1 5, the corresponding relationship between the level of throttling required to adjust the throttling degree of the processor and the temperature difference between the processor temperature and the system temperature (not shown in the figure) must be set. The adjustment method can be: when the detected temperature difference between the processor temperature and the system temperature is within a predetermined value range, the throttle degree of the processor is fixed at a predetermined level corresponding to the predetermined value range . Please refer to Figure 4 which shows the relationship between processor utilization and temperature under ideal conditions. As can be seen from the figure, within a specific rated temperature, the processor utilization is directly proportional to the temperature. This is because when the processor is running

11597t.wf.ptd Page 10 1226530 V. Explanation of the invention (6) When the speed is fast, the temperature of the processor itself will increase with the increase of the processor usage rate. This is because the processor works Clocks increase. However, once the operating temperature of the processor is exceeded, this operating temperature is usually the highest allowable temperature at which the processor can maintain normal operation. The throttling function of the processor is enabled to reduce the processor utilization rate and protect the processor from Overheated and damaged. However, in practical applications, it is extremely unlikely to achieve such an ideal relationship. Next, please refer to FIG. 5 which is a schematic diagram showing the relationship between the processor usage rate and the temperature according to an embodiment of the present invention. In this embodiment, although the processor usage rate is divided into 12.5%, it is divided into 12.5%, 25%, 37 · 5%, 50%, 62,5%, 75%, 87.5 %, 100%, etc., but there is no need to limit it in practical applications. Although the present invention only discloses the above embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications and retouches without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention It shall be subject to the definition in the appended patent application scope.

11597twf.ptd Page 11 1226530 Brief description of the diagrams Figure 1 is a schematic diagram showing the relationship between processor utilization and temperature of the conventional technology. FIG. 2 is a schematic flowchart illustrating a first embodiment according to the present invention. FIG. 3 is a schematic flowchart illustrating a second embodiment according to the present invention. Figure 4 shows the relationship between processor utilization and temperature under ideal conditions. FIG. 5 is a schematic diagram showing the relationship between processor utilization and temperature according to an embodiment of the present invention. Symbol description: S203 ~ S212: Implementation steps S3 0 3 ~ S 3 1 5 •• Implementation steps

11597twf. Pt.d Page 12

Claims (1)

1226530 VI. Scope of Patent Application1. A method for reducing the energy consumption of a processor, which is suitable for dividing the throttling degree of a processor into multiple levels. The method for reducing the energy consumption of a processor includes: When the processor is turned on, the throttling degree of the processor is set to any of the layers except the layer with the highest usage rate of the processor; detecting the temperature of the processor; and the processor according to the detection The temperature of the processor is adjusted accordingly. 2. The method for reducing energy consumption of a processor as described in item 1 of the scope of patent application, wherein the adjustment of the throttling degree of the processor includes when the detected temperature of the processor is within a predetermined value range, Then, the throttling degree of the processor is set to a predetermined level corresponding to the predetermined value range. 3. The method for reducing energy consumption of a processor as described in item 1 of the scope of patent application, wherein the degree of throttling of the processor is divided into eight levels. 4. The method for reducing energy consumption of a processor as described in item 1 of the scope of patent application, wherein before adjusting the throttling degree of the processor according to the detected temperature of the processor, the method further includes: The correspondence relationship between the throttling degree hierarchy required to adjust the throttling degree of the processor and the processor temperature is set. 5, — A method for reducing the energy consumption of a processor, which is suitable for dividing the throttling degree of a processor into a plurality of levels. The method for reducing the energy consumption of a processor includes: when the processor is turned on, Set the throttling degree of the processor to any of the layers except the one with the highest utilization of the processor; 11597t.wf.ptd Page 13 1226530 6. Apply for a patent to detect the temperature of the processor; detect the temperature of the system where the processor is located; and based on the difference between the temperature of the processor and the temperature of the system, Adjust the throttling degree of the processor accordingly. 6. The method of reducing energy consumption of a processor as described in item 5 of the scope of patent application, wherein the adjustment of the throttling degree of the processor includes setting the processor's The degree of throttling is at a predetermined level corresponding to the predetermined value range. 7. The method for reducing energy consumption of a processor as described in item 5 of the scope of patent application, wherein the degree of throttling of the processor is divided into eight levels. 8 β The method for reducing energy consumption of a processor as described in item 5 of the scope of patent application, wherein before adjusting the throttling degree of the processor correspondingly according to the difference, the method further includes: The corresponding relationship between the level of throttling required by the throttling degree and the difference between the processor temperature and the system temperature. 11597twf.ptd Page 14