A multi-modal fan control strategy method based on real-time power consumption perception

Abstract

The application relates to a multi-modal fan control strategy method based on real-time power consumption perception, which comprises the following steps: step 1, multi-source heterogeneous power consumption perception and data normalization are established to provide standardized real-time power data for upper-layer control logic; step 2, a deterministic finite state machine based on real-time average power driven by power consumption perception is established; and step 3, an asymmetric hysteresis mechanism in the dual dimensions of power amplitude and duration is established, and corresponding operations are made according to the mechanism. The application improves the mute performance and acoustic quality under all working conditions.

Description

Technical Field

[0001] This invention relates to the field of thermal management of electronic devices, and in particular to a multimodal fan control strategy method based on real-time power consumption awareness. Background Technology

[0002] With the evolution of semiconductor technology and the rapid improvement of computing performance, the heat flux density of high-performance computing chips (such as processors, graphics processing units, etc.) continues to surge, posing a severe challenge to the heat dissipation management of electronic devices.

[0003] Current fan speed control logic generally relies on feedback signals from temperature sensors. However, from the generation of core heat to the detection by the temperature sensor, multiple physical media, including chip packaging, thermal interface materials, and heat sink base, must pass through. The temperature signal exhibits a significant physical phase lag relative to real-time power consumption. This means that when the chip load surges instantaneously, the fan often cannot respond in time, leading to transient thermal shocks to the chip core. This can easily trigger hardware-level thermal throttling, severely inhibiting performance release. When the sensor detects high temperatures, heat has already accumulated in the cooling system, and the fan must intervene at extremely high speeds (usually accompanied by loud noise) to suppress the temperature rise, resulting in a "fan roar."

[0004] A single temperature curve essentially projects a complex state space containing multidimensional variables such as power consumption, ambient temperature, and transient load onto a one-dimensional linear "temperature-speed" axis. Since the relationship between power consumption and temperature rise is not a simple linear one, a one-dimensional temperature curve cannot distinguish the essential difference between "high ambient temperature" and "high load heat." This information loss causes control commands to often exhibit distortion and lag when the system handles dynamic loads, making it difficult to achieve optimal energy efficiency.

[0005] In real-world operating conditions, processor frequency scheduling typically manifests as high-frequency pulses, leading to drastic fluctuations in temperature signals. One-dimensional curves based on temperature feedback are highly susceptible to "control oscillations" or "ping-pong effects" at critical thresholds. This not only causes frequent and abrupt fluctuations in fan speed, severely interfering with the user's subjective acoustic experience, but also shortens the mechanical lifespan of the cooling fan. Summary of the Invention

[0006] To address the aforementioned technical problems, the purpose of this invention is to provide a multimodal fan control strategy method based on real-time power consumption awareness.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: A multimodal fan control strategy method based on real-time power consumption awareness includes the following steps: Step 1: Establish multi-source heterogeneous power consumption sensing and data normalization to provide standardized real-time power data for upper-level control logic; Step 11: Establish a hardware abstraction layer to achieve adaptive hardware recognition; Step 12: Establish a data normalization layer to convert raw data in different formats into a unified physical quantity, watts. Step 2: Establish a deterministic finite state machine driven by real-time average power based on power consumption awareness; Step 21: Establish energy efficiency status classification; Step 22: Establish independent binding for curves; Step 23: Establish feedforward control logic; Step 3: Establish an asymmetric hysteresis mechanism with two dimensions: power amplitude and duration, and perform corresponding operations based on the mechanism.

[0008] Preferably, in the multimodal fan control strategy method based on real-time power consumption awareness, step 11 establishes a hardware abstraction layer for different architectures, including Intel architecture and heterogeneous processing platforms.

[0009] Preferably, the multimodal fan control strategy method based on real-time power consumption awareness is designed for Intel architecture: the embedded controller (EC) reads the "Accumulated Energy Status" and "Package Power SKU Unit" through the PECI bus and performs power consumption calculation.

[0010] Preferably, in the aforementioned multimodal fan control strategy method based on real-time power consumption awareness, for the digital voltage regulator (VRM) in the heterogeneous processing platform: the embedder controller (EC) directly accesses the power register through the I²C interface and executes the data format decoding logic.

[0011] Preferably, in the heterogeneous processing platform, for analog VRM, the method of multimodal fan control strategy based on real-time power consumption awareness is: the pins are sampled by ADC through the embedder controller (EC) to calculate the real-time average power.

[0012] Preferably, the multimodal fan control strategy method based on real-time power consumption awareness classifies energy efficiency states according to real-time average power (P). avg The operating state is divided into three discrete states: silent state (S0), balanced state (S1), and performance state (S2).

[0013] Preferably, the multimodal fan control strategy method based on real-time power consumption awareness has independently bound curves, and each state is equipped with a corresponding power envelope interval depth and an independent fan curve, achieving an ultimate balance between fan noise and processor performance.

[0014] Preferably, the multimodal fan control strategy method based on real-time power consumption awareness includes a feedforward control logic that, when power consumption is detected to cross the state boundary, the system directly switches to the corresponding fan curve, thereby implementing a "pre-cooling" strategy before the core temperature actually rises, suppressing transient thermal shock.

[0015] Preferably, in the multimodal fan control strategy method based on real-time power consumption awareness, the mechanism in step 3 includes: Upgrade triggering: Utilizing the leading indicator characteristic of power consumption signal relative to temperature rise response, such as real-time average power (P). avg ) meets the switching baseline threshold (P) th ), that is, satisfying the decision logic (P) avg >P th When the temperature rise slope steepens, the system immediately executes a "preemptive response" to build a pre-cooling environment in advance and suppress transient thermal shock. Downshifting constraint: When exiting high-power mode, both amplitude and time domain conditions must be met simultaneously, and the power must be within a preset time window. The internal stability must be maintained, and the real-time power must fall back to the level including the hysteresis margin P. hys The lower limit threshold (P) avg <(P th -P hys By compensating for the heat dissipation process of the cooling system, it ensures that the residual heat is fully discharged, filters the fan "surge" noise caused by transient load fluctuations, and improves the stability of acoustic quality.

[0016] Preferably, in the multimodal fan control strategy method based on real-time power consumption awareness, the power consumption is within a preset time window. The duration of stability is 3-5 seconds.

[0017] By means of the above-described solution, the present invention has at least the following advantages: This invention abandons the single fan curve and establishes a multimodal energy efficiency state machine (quiet state, balanced state, performance state) based on average power consumption mapping. Each state is equipped with an independent fan curve that is deeply optimized for the power envelope range, so that the system can accurately eliminate heat dissipation redundancy and limit the fan speed to the minimum acoustic threshold required to maintain thermal balance. This fundamentally smooths the speed rise slope and improves the quiet performance and acoustic quality under all operating conditions.

[0018] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the principle of the present invention; Figure 2 It is a flowchart of the invention based on a finite state machine driven by real-time average power; Figure 3 It is an asymmetric hysteresis timing diagram based on power amplitude and time domain. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0022] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0023] Example like Figures 1 to 3 As shown, a multimodal fan control strategy method based on real-time power consumption awareness includes the following steps: Step 1: Establish multi-source heterogeneous power consumption sensing and data normalization to provide standardized real-time power data for upper-level control logic; Step 11: Establish a hardware abstraction layer to achieve adaptive hardware recognition; Identification and processing for Intel architecture and heterogeneous processing platforms: For Intel architecture: The Embedded Controller (EC) reads the "Accumulated Energy Status" and "Package Power SKU Unit" via the PECI bus and performs power consumption calculations; In heterogeneous processing platforms for digital voltage regulators (VRMs): the embedded controller (EC) directly accesses the power register via the I²C interface and performs data format decoding logic.

[0024] For analog VRMs in heterogeneous processing platforms: the pins are sampled by an ADC through an embedder controller (EC) to calculate the real-time average power.

[0025] Step 12: Establish a data normalization layer to convert raw data in different formats into a unified physical quantity, watts. Step 2: Establish a deterministic finite state machine driven by real-time average power based on power consumption awareness; Step 21: Establish energy efficiency state classification based on real-time average power (P). avg The operating state is divided into three discrete states: silent state (S0), balanced state (S1), and performance state (S2). Step 22: Establish independent curve binding; each state is equipped with a specific power envelope depth and an independent fan curve, achieving an ultimate balance between fan noise and processor performance; Step 23: Establish feedforward control logic. When the power consumption is detected to cross the state boundary, the system directly switches to the corresponding fan curve, thereby implementing a "pre-cooling" strategy before the core temperature actually rises to suppress transient thermal shock. Step 3: Establish an asymmetric hysteresis mechanism with two dimensions: power amplitude and duration, and perform corresponding operations based on the mechanism.

[0026] The mechanism in step 3 includes: Upgrade triggering: Utilizing the leading indicator characteristic of power consumption signal relative to temperature rise response, such as real-time average power (P). avg ) meets the switching baseline threshold (P) th ), that is, satisfying the decision logic (P) avg >P th When the temperature rise slope steepens, the system immediately executes a "preemptive response" to build a pre-cooling environment in advance and suppress transient thermal shock. Downshifting constraint: When exiting high-power mode, both amplitude and time domain conditions must be met simultaneously, and the power must be within a preset time window. The voltage must remain stable for 3-5 seconds, and the real-time power must drop back to the level including the hysteresis margin P. hys The lower limit threshold (P) avg <(P th -P hys By compensating for the heat dissipation process of the cooling system, it ensures that the residual heat is fully discharged, filters the fan "surge" noise caused by transient load fluctuations, and improves the stability of acoustic quality.

[0027] Among them, Figure 3 The top figure shows the relationship between real-time average power and hysteresis threshold; the middle figure shows the gear switching of the multi-modal fan control strategy; and the bottom figure shows the timing diagram of the fan PWM duty cycle response.

[0028] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0029] In the description of this application, it should be noted that the terms "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product is in use. They are used only for the convenience of describing this application and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0030] Furthermore, terms such as "horizontal" and "vertical" do not imply that components must be absolutely horizontal or vertical, but rather that they can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0031] In the description of this application, it should also be noted that, unless otherwise expressly specified and limited, the terms "set up," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0032] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A multimodal fan control strategy method based on real-time power consumption awareness. Its features are, Includes the following steps: Step 1: Establish multi-source heterogeneous power consumption sensing and data normalization to provide standardized real-time power data for upper-level control logic; Step 11: Establish a hardware abstraction layer to achieve adaptive hardware recognition; Step 12: Establish a data normalization layer to convert raw data in different formats into a unified physical quantity, watts. Step 2: Establish a deterministic finite state machine driven by real-time average power based on power consumption awareness; Step 21: Establish energy efficiency status classification; Step 22: Establish independent binding for curves; Step 23: Establish feedforward control logic; Step 3: Establish an asymmetric hysteresis mechanism with two dimensions: power amplitude and duration, and perform corresponding operations based on the mechanism.

2. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 1, characterized in that: Step 11 establishes a hardware abstraction layer for different architectures, including Intel architecture and heterogeneous processing platforms.

3. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 2, characterized in that: For Intel architecture: The Embedded Controller (EC) reads the "Accumulated Energy Status" and "Package Power SKU Unit" via the PECI bus and performs power consumption calculations.

4. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 2, characterized in that: In heterogeneous processing platforms for digital voltage regulators (VRMs): the embedded controller (EC) directly accesses the power register via the I²C interface and performs data format decoding logic.

5. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 2, characterized in that: For analog VRMs in heterogeneous processing platforms: the pins are sampled by an ADC through an embedder controller (EC) to calculate the real-time average power.

6. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 1, characterized in that: Energy efficiency status classification is based on real-time average power (P). avg The operating state is divided into three discrete states: silent state (S0), balanced state (S1), and performance state (S2).

7. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 1, characterized in that: The curves are independently bound, with a specific power envelope depth and an independent fan curve for each state.

8. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 1, characterized in that: The feedforward control logic detects that the power consumption exceeds the state boundary, and the system directly switches to the corresponding fan curve, thereby implementing a "pre-cooling" strategy before the core temperature actually rises, suppressing transient thermal shock.

9. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 1, characterized in that: The mechanism in step 3 includes: Upgrade triggering: Utilizing the leading indicator characteristic of power consumption signal relative to temperature rise response, such as real-time average power (P). avg ) meets the switching baseline threshold (P) th ), that is, satisfying the decision logic (P) avg >P th When the temperature rise slope steepens, the system immediately executes a "preemptive response" to build a pre-cooling environment in advance and suppress transient thermal shock. Downshifting constraint: When exiting high-power mode, both amplitude and time domain conditions must be met simultaneously, and the power must be within a preset time window. The internal stability must be maintained, and the real-time power must fall back to the level including the hysteresis margin P. hys The lower limit threshold (P) avg <(P th -P hys By compensating for the heat dissipation process of the cooling system, it ensures that excess heat is fully discharged, filters out fan noise caused by transient load fluctuations, and improves the stability of acoustic quality.

10. The multimodal fan control strategy method based on real-time power consumption awareness according to claim 9, characterized in that: Power within the preset time window The duration of stability is 3-5 seconds.

Images