Energy efficiency optimization method of network interface equipment, equipment, medium and product

Abstract

The invention provides an energy efficiency optimization method of network interface equipment, equipment, a medium and a product, and relates to the technical field of network interface equipment. The method comprises the following steps: determining a predicted energy efficiency ratio of the network interface equipment according to real-time performance parameters of the network interface equipment; if it is detected that the network interface equipment is in an energy efficiency depression state according to the real-time performance parameters and the predicted energy efficiency ratio, determining a target voltage and a target frequency according to the real-time performance parameters and the predicted energy efficiency ratio, and adjusting the voltage and the frequency of the network interface equipment by adopting a dynamic working point jumping mechanism; if it is detected that the network interface device is not in the energy efficiency depression state and the voltage frequency adjustment mode is determined to be boost frequency increasing or buck frequency decreasing according to the real-time performance parameters and the predicted energy efficiency ratio, the voltage and frequency of the network interface device are fed back and adjusted according to the voltage frequency adjustment mode until it is determined that the voltage frequency adjustment mode is adjustment stopping. According to the method provided by the invention, linearized stable output of the energy efficiency ratio of the network interface equipment is realized.

Description

Technical Field

[0001] The present application relates to the technical field of network interface devices, and in particular to a method, device, medium and product for optimizing the energy efficiency of a network interface device. Background Art

[0002] The network interface device is an infrastructure computing platform designed for software-defined networking, storage acceleration, and network security tasks. It is suitable for various application scenarios such as accelerating artificial intelligence (AI), hybrid cloud, high-performance computing, and fifth-generation mobile communication technology (5G) wireless networks.

[0003] In related technologies, network interface devices are dynamically switched in connection state (e.g., active or idle) based on their bandwidth utilization to achieve power consumption mode switching. However, this approach ignores the fact that network interface devices may be in energy efficiency troughs and, in turn, the nonlinear energy efficiency characteristics of the medium bandwidth utilization range. As a result, related technologies have difficulty achieving a linear and stable output of the energy efficiency ratio of network interface devices. Summary of the Invention

[0004] The present application provides a method, device, medium and product for optimizing the energy efficiency of a network interface device, which realizes the linear and stable output of the energy efficiency ratio of the network interface device.

[0005] This application provides a method for optimizing energy efficiency of a network interface device, comprising:

[0006] Obtain real-time performance parameters of network interface devices;

[0007] determining a predicted energy efficiency ratio of the network interface device based on real-time performance parameters of the network interface device;

[0008] Detect whether the network interface device is in an energy efficiency trough state based on real-time performance parameters and predicted energy efficiency ratio;

[0009] If it is detected that the network interface device is in an energy efficiency trough state, a target voltage and a target frequency are determined based on the real-time performance parameters and the predicted energy efficiency ratio; wherein the target voltage and the target frequency are the voltage and the frequency when the network interface device is out of the energy efficiency trough state;

[0010] Adopting dynamic operating point jump mechanism to adjust the voltage and frequency of network interface devices according to target voltage and target frequency;

[0011] If it is detected that the network interface device is not in an energy efficiency trough state, the voltage and frequency adjustment method is determined based on the real-time performance parameters and the predicted energy efficiency ratio;

[0012] If it is determined that the voltage and frequency adjustment mode is voltage and frequency increase or voltage and frequency reduction, the voltage and frequency of the network interface device are adjusted according to the voltage and frequency adjustment mode until it is determined that the voltage and frequency adjustment mode is to stop adjustment.

[0013] The present application also provides an energy efficiency optimization device for a network interface device, comprising:

[0014] A first acquisition module is used to obtain real-time performance parameters of the network interface device;

[0015] An energy efficiency ratio calculation module, used to determine a predicted energy efficiency ratio of the network interface device based on real-time performance parameters of the network interface device;

[0016] A depression detection module is used to detect whether the network interface device is in an energy efficiency depression state based on real-time performance parameters and predicted energy efficiency ratio;

[0017] a target value calculation module, configured to determine a target voltage and a target frequency based on real-time performance parameters and a predicted energy efficiency ratio if the network interface device is detected to be in an energy efficiency trough state; wherein the target voltage and the target frequency are the voltage and the frequency when the network interface device exits the energy efficiency trough state;

[0018] A jump adjustment module, configured to adjust the voltage and frequency of the network interface device according to a target voltage and a target frequency by adopting a dynamic operating point jump mechanism;

[0019] An adjustment determination module, configured to determine a voltage and frequency adjustment method based on real-time performance parameters and a predicted energy efficiency ratio if it is detected that the network interface device is not in an energy efficiency trough state;

[0020] The adaptive adjustment module is used to feedback adjust the voltage and frequency of the network interface device according to the voltage and frequency adjustment mode if it is determined that the voltage and frequency adjustment mode is boosting and increasing the frequency or reducing the voltage and frequency, until it is determined that the voltage and frequency adjustment mode is stopping the adjustment.

[0021] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of any of the above-mentioned methods for optimizing the energy efficiency of a network interface device when executing the computer program.

[0022] The present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned methods for optimizing the energy efficiency of a network interface device are implemented.

[0023] The present application also provides a computer program product, including a computer program, which implements the steps of any of the above-mentioned methods for optimizing the energy efficiency of a network interface device when the computer program is executed by a processor.

[0024] The present application provides a method, device, medium, and product for optimizing the energy efficiency of a network interface device. The method includes: determining a predicted energy efficiency ratio of the network interface device based on real-time performance parameters of the network interface device; if the network interface device is detected to be in an energy efficiency trough state based on the real-time performance parameters and the predicted energy efficiency ratio, determining a target voltage and a target frequency based on the real-time performance parameters and the predicted energy efficiency ratio, and adjusting the voltage and frequency of the network interface device using a dynamic operating point jump mechanism; if it is detected that the network interface device is not in an energy efficiency trough state, and if the voltage and frequency adjustment mode is determined to be boost and increase or buck and decrease and decrease based on the real-time performance parameters and the predicted energy efficiency ratio, feedback adjustment of the voltage and frequency of the network interface device is performed based on the voltage and frequency adjustment mode until it is determined that the voltage and frequency adjustment mode is to stop adjustment. The following technical effects are achieved: through continuous prediction, detection and feedback adjustment, the energy efficiency fluctuations of network interface devices under different loads are controlled, and the linear and stable output of the energy efficiency ratio of the network interface devices is achieved; according to real-time performance parameters and predicted energy efficiency ratios, it is detected whether the network interface device is in an energy efficiency trough state, and through a dynamic working point jump mechanism, the voltage and frequency of the network interface device are adjusted, thereby achieving active detection and avoidance of the energy efficiency trough state, and thus optimizing the nonlinear energy efficiency in the medium load area. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0026] Figure 1 Schematic diagram of the process of the energy efficiency optimization method of the network interface device provided in the embodiment of the present application Figure 1 ;

[0027] Figure 2 Schematic diagram of the process of the energy efficiency optimization method of the network interface device provided in the embodiment of the present application Figure 2 ;

[0028] Figure 3 Schematic diagram of the process of the energy efficiency optimization method of the network interface device provided in the embodiment of the present application Figure 3 ;

[0029] Figure 4 A schematic diagram of the structure of an energy efficiency optimization device for a network interface device provided in an embodiment of the present application;

[0030] Figure 5 This is a schematic diagram of the structure of the electronic device provided in this application.

[0031] Reference numerals:

[0032] 410 - first acquisition module; 420 - energy efficiency ratio calculation module; 430 - depression detection module; 440 - target value calculation module; 450 - jump adjustment module; 460 - adjustment determination module; 470 - adaptive adjustment module;

[0033] 510 - processor; 520 - memory; 530 - communication component. DETAILED DESCRIPTION

[0034] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0035] It should be noted that in the description of this application, the terms "include", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. The terms "first", "second" and the like in this application are used to distinguish similar objects, and are not used to describe a specific order or sequence. The user information (including but not limited to user device information and user personal information, etc.) and data (including but not limited to data for analysis, stored data and displayed data, etc.) involved in one or more embodiments of this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant laws, regulations and standards, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0036] In order to clearly understand the technical solution of the present application, the solution of the prior art is first introduced in detail.

[0037] The network interface device is an infrastructure computing platform designed for software-defined networking, storage acceleration, and network security tasks. It is suitable for various application scenarios such as accelerating artificial intelligence (AI), hybrid cloud, high-performance computing, and fifth-generation mobile communication technology (5G) wireless networks.

[0038] In related technologies, network interface devices are dynamically switched in connection state (e.g., active or idle) based on their bandwidth utilization to achieve power consumption mode switching. However, this approach ignores the fact that network interface devices may be in energy efficiency troughs, and thus neglects the nonlinear energy efficiency characteristics in the medium bandwidth utilization range.

[0039] Furthermore, related technologies rely on software interrupts to switch power consumption modes, resulting in high response delays and an inability to cope with microsecond-level traffic bursts.

[0040] Furthermore, related technologies use the Autoregressive Integrated Moving Average (ARIMA) algorithm in conjunction with Dynamic Voltage and Frequency Scaling (DVFS) to optimize energy efficiency. Specifically, this algorithm uses a fixed multi-level voltage switching mechanism, such as 0.8V, 0.9V, 1.0V, 1.1V, and 1.2V, to adjust the voltage and frequency ratio. However, this algorithm suffers from slow prediction speed and limited flexibility in the fixed voltage levels.

[0041] In summary, it is difficult for related technologies to achieve linear and stable output of the energy efficiency ratio of network interface devices.

[0042] Therefore, in response to the above technical issues, the research found that in order to solve the problem, when the network interface device is detected to be in an energy efficiency trough state, the inefficient working point is intelligently skipped; otherwise, the voltage and frequency of the network interface device are independently adjusted through multi-voltage domain fast switching and dual-clock architecture to break through the linear constraints of DVFS; at the same time, a high-precision energy efficiency model is established through nonlinear multi-order polynomials to achieve dynamic prediction of the energy efficiency ratio.

[0043] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0044] In conjunction with the specific application environment architecture or specific hardware architecture on which the execution of the energy efficiency optimization method of the network interface device depends, the specific application environment architecture or specific hardware architecture is described herein.

[0045] The core of the energy efficiency optimization method for network interface devices provided in the embodiments of this application lies in a server equipped with the network interface device. This method is widely used in cutting-edge application scenarios with high network energy efficiency requirements. The network interface device can be a Smart NIC supporting Adaptive Voltage and Frequency Scaling (AVFS), a Field Programmable Gate Array (FPGA), or an Application-Specific Integrated Circuit (ASIC); the server can be a data center server, an edge computing node, a high-performance computing cluster, or a virtualization platform.

[0046] This method is suitable for accelerating artificial intelligence scenarios. Specifically, in AI training and inference clusters, large amounts of model parameters and data need to be frequently exchanged between servers. These exchanged data are characterized by strong bursts and high bandwidth requirements. Servers equipped with network interface devices can predict burst traffic in real time and, based on current temperature and packet characteristics, precisely adjust the voltage and frequency of the network interface devices to ensure sufficient performance during peak AI task times and quickly enter an efficient energy-saving state during low times, avoiding energy efficiency troughs at medium loads, thereby significantly reducing the total cost of ownership of the entire AI cluster.

[0047] This approach is also applicable to hybrid cloud scenarios. Specifically, hybrid cloud environments involve complex data migration, synchronization, and security policy enforcement between public and private clouds. Network interface devices (NIDs) fulfill the important role of software-defined network storage and security acceleration. Servers equipped with NIDs utilize nonlinear multi-order polynomial modeling to accurately adapt to the variable and unpredictable network load patterns in hybrid cloud scenarios, achieving linear and stable output of energy efficiency ratios, ensuring the stability and efficiency of cross-cloud services while optimizing cloud resource energy consumption.

[0048] This approach is also applicable to high-performance computing (HPC) scenarios. Specifically, HPC applications typically require extremely large-scale parallel computing, where inter-node communication becomes a performance bottleneck. The line-speed processing capabilities of network interface devices are crucial. Servers equipped with network interface devices leverage their low latency and high precision to dynamically adjust the power consumption of the network interface devices, ensuring that they always operate at optimal energy efficiency during long, high-intensity computing tasks, effectively controlling data center cooling pressure and operating costs.

[0049] This approach is also applicable to 5G wireless network scenarios. Specifically, 5G core networks and edge computing nodes must handle massive connection requests and data flows from mobile devices, resulting in highly dynamic traffic patterns. Servers equipped with network interface devices can rapidly respond to microsecond-level traffic bursts and effectively suppress temperature drift, ensuring high reliability and low-latency services in 5G networks while improving the energy efficiency of network infrastructure.

[0050] Figure 1 Schematic diagram of the process of the energy efficiency optimization method of the network interface device provided in the embodiment of the present application Figure 1 .like Figure 1 As shown, in the embodiment of the present application, the execution subject may be an energy efficiency optimization device for a network interface device, which may be located in an electronic device, which may be a server equipped with a network interface device. The energy efficiency optimization method for a network interface device provided in the embodiment of the present application includes the following steps:

[0051] S101. Acquire real-time performance parameters of a network interface device.

[0052] Specifically, real-time performance parameters refer to parameters across multiple dimensions that can instantly reflect the current operating status of network interface devices. These parameters can include key indicators such as normalized bandwidth utilization, operating temperature, average packet length, queue depth, and normalized processor utilization.

[0053] S102: Determine a predicted energy efficiency ratio of the network interface device according to real-time performance parameters of the network interface device.

[0054] Specifically, the predicted energy efficiency ratio (PEER) refers to the energy efficiency level that a network interface device is expected to achieve under current real-time performance parameters, reflecting its processing power per unit of power consumption. Servers can determine the predicted EER of a network interface device using machine learning models such as lightweight neural networks or random forests (which are pre-trained using large amounts of historical operating data), or using pre-configured nonlinear multi-order polynomials.

[0055] S105: Detect whether the network interface device is in an energy efficiency trough state based on the real-time performance parameters and the predicted energy efficiency ratio.

[0056] Specifically, an energy efficiency trough occurs when a network interface device operates within a specific performance parameter threshold range, but its actual energy efficiency ratio is significantly lower than the theoretical optimal value, exhibiting a significant decrease in energy efficiency. Operating a server in this state will result in increased energy consumption and decreased performance. Based on real-time performance parameters, the server determines whether the current real-time performance parameters fall within the performance parameter threshold range, and combines this with whether the predicted energy efficiency ratio is significantly lower than the expected value under the real-time performance parameters to comprehensively determine whether the network interface device is in an energy efficiency trough.

[0057] If it is detected that the network interface device is in an energy efficiency low state, then the process continues with S106 ; if it is detected that the network interface device is not in an energy efficiency low state, then the process continues with S108 .

[0058] S106: Determine the target voltage and target frequency according to the real-time performance parameters and the predicted energy efficiency ratio.

[0059] Specifically, the target voltage and target frequency are calculated specifically to quickly remove the network interface device from an energy-efficiency trough. These target voltages and frequencies are the voltages and frequencies required to remove the device from an energy-efficiency trough. Once the server detects that the device is in an energy-efficiency trough, it uses an energy-efficiency gradient optimization algorithm based on current real-time performance parameters and predicted energy efficiency ratios to accurately calculate the target voltages and frequencies required to remove the device from an energy-efficiency trough.

[0060] S107 , using a dynamic operating point jump mechanism to adjust the voltage and frequency of the network interface device according to the target voltage and target frequency.

[0061] Specifically, the dynamic operating point jump mechanism is a voltage and frequency adjustment strategy. When the server detects that a network interface device is experiencing energy inefficiency, it quickly adjusts the voltage and frequency to pre-calculated target values, effectively avoiding these inefficient areas. Voltage and frequency adjustments can be performed directly to the target value in a single step, or using multi-step incremental jumps to quickly reach the target value in multiple steps, further reducing the risk of transient current surges or clock instability caused by sudden voltage or frequency changes.

[0062] S108. Determine a voltage and frequency adjustment method based on the real-time performance parameters and the predicted energy efficiency ratio.

[0063] Specifically, the voltage and frequency adjustment method determines the direction of voltage and frequency adjustment based on real-time performance parameters and predicted energy efficiency. This adjustment method includes three modes: boost and frequency, buck and frequency, and stop adjustment. Boost and frequency refers to increasing voltage and frequency to enhance performance when the load increases; buck and frequency refers to reducing voltage and frequency to conserve power when the load decreases; and stop adjustment indicates that the current voltage and frequency have reached or are near optimal and no further adjustment is required.

[0064] If it is determined that the voltage-frequency adjustment mode is voltage-up and frequency-up or voltage-down and frequency-down, then the process continues with S109 .

[0065] S109 , feedback-adjusting the voltage and frequency of the network interface device according to the voltage and frequency adjustment mode, until determining that the voltage and frequency adjustment mode is to stop adjustment.

[0066] Specifically, if a voltage and frequency increase or decrease is determined, the server will gradually adjust the voltage and frequency accordingly, continuously monitoring the effect of the adjustment. This process is a feedback mechanism until the server determines that the current voltage and frequency have reached or are close to the optimal state, at which point it confirms that voltage and frequency adjustments will cease.

[0067] The voltage and frequency can be adjusted using adaptive voltage and frequency scaling (AVFS), which dynamically adjusts the voltage and frequency based on real-time performance parameters to meet performance requirements while minimizing power consumption.

[0068] The voltage and frequency can also be adjusted using proportional-integral-derivative (PID) control. By analyzing the energy efficiency deviation, its cumulative amount, and rate of change, the voltage / frequency adjustment is calculated using a linear combination of proportional, integral, and differential to achieve stable and precise feedback control.

[0069] The voltage and frequency can also be adjusted using fuzzy logic control (Fuzzy Logic Control). This is based on the fuzzy state of inputs such as energy efficiency deviation and temperature and preset expert rules, and performs reasoning and defuzzification to derive nonlinear voltage / frequency adjustment instructions.

[0070] The voltage and frequency can also be adjusted using a lookup table (LUT)-based adaptive step size adjustment. This involves retrieving the corresponding voltage / frequency adaptive step size from a preset multi-dimensional lookup table based on real-time performance parameters to achieve fast, offline optimized feedback adjustment.

[0071] Voltage and frequency can also be adjusted online using reinforcement learning (RL). This approach treats voltage / frequency adjustment as a decision-making action, obtains reward signals such as energy efficiency improvement through interaction with network interface devices, and learns the optimal strategy online to maximize long-term energy efficiency.

[0072] After executing S107 or S109, or after executing S108, if the voltage and frequency adjustment mode is determined to be stopped, the server will return to the starting point of the loop, re-executing S101 to optimize the energy efficiency of the network interface device for the next time period, thereby achieving continuous closed-loop control. At the same time, the voltage and frequency parameters of the network interface device after this adjustment will directly affect the operating mode of the next loop, determining whether to execute dynamic jump decisions again or perform adaptive energy efficiency fine-tuning. This mechanism ensures that the network interface device can always operate stably at or near-optimal energy efficiency, achieving a linear and stable output of its energy efficiency ratio.

[0073] An embodiment of the present application provides an energy efficiency optimization method for a network interface device, comprising: determining a predicted energy efficiency ratio of the network interface device based on real-time performance parameters of the network interface device; if the network interface device is detected to be in an energy efficiency trough state based on the real-time performance parameters and the predicted energy efficiency ratio, determining a target voltage and a target frequency based on the real-time performance parameters and the predicted energy efficiency ratio, and adjusting the voltage and frequency of the network interface device using a dynamic operating point jump mechanism; if it is detected that the network interface device is not in an energy efficiency trough state, and if the voltage and frequency adjustment mode is determined to be boost and increase or buck and decrease based on the real-time performance parameters and the predicted energy efficiency ratio, feedback adjustment of the voltage and frequency of the network interface device is performed based on the voltage and frequency adjustment mode until it is determined that the voltage and frequency adjustment mode is to stop adjustment. The following technical effects are achieved: through continuous prediction, detection and feedback adjustment, the energy efficiency fluctuations of network interface devices under different loads are controlled, and the linear and stable output of the energy efficiency ratio of the network interface devices is achieved; according to real-time performance parameters and predicted energy efficiency ratios, it is detected whether the network interface device is in an energy efficiency trough state, and through a dynamic working point jump mechanism, the voltage and frequency of the network interface device are adjusted, thereby achieving active detection and avoidance of the energy efficiency trough state, and thus optimizing the nonlinear energy efficiency in the medium load area.

[0074] Figure 2 Schematic diagram of the process of the energy efficiency optimization method of the network interface device provided in the embodiment of the present application Figure 2 .like Figure 2 As shown, in one possible design, S109, wherein feedback adjustment of the voltage of the network interface device according to the voltage-frequency adjustment method includes:

[0075] S201 : Determine an adjusted target voltage domain from a plurality of preset voltage domains according to a voltage-frequency adjustment method.

[0076] Among them, each voltage domain corresponds to an independent power supply network.

[0077] If it is detected that the current original voltage domain is inconsistent with the target voltage domain, then continue to execute S202; otherwise, end the adjustment.

[0078] S202: Switch to the power supply network corresponding to the target voltage domain for power supply.

[0079] Specifically, a power network refers to the physically isolated circuit paths that provide power to specific voltage domains. These include independent power lines, voltage regulators, and control logic, ensuring electrical isolation between different voltage domains to prevent mutual interference. Voltage domains divide the power supply system of network interface devices into multiple independent power supply areas with varying voltage levels. Each voltage domain can independently enable, disable, or adjust its output voltage.

[0080] Based on the determined voltage frequency adjustment method, the server selects a voltage domain that matches the required voltage level from multiple preset voltage domains as the target voltage domain. For example, if a voltage boost is required, the next voltage domain with a higher voltage level than the current voltage domain is selected; conversely, if a voltage reduction is required, the next voltage domain with a lower voltage level than the current voltage domain is selected. After determining the target voltage domain, the server checks whether the current voltage domain and the target voltage domain are consistent. If they are inconsistent, the server will initiate a power switching process, switching the power supply source of the network interface device from the power supply network corresponding to the current voltage domain to the power supply network corresponding to the target voltage domain, with the latter providing power.

[0081] The technical effect of the embodiments of the present application is that by dividing the power supply system into multiple independent voltage domains, independent and dynamic voltage adjustment requirements are achieved.

[0082] In one possible design, the step of switching to a power supply network corresponding to the target voltage domain for power supply in S202 includes:

[0083] S2021: Pre-charge the power supply network corresponding to the target voltage domain until the pre-set first time duration is reached.

[0084] S2022: Switch to the power supply network corresponding to the target voltage domain for power supply, and disconnect the power supply connection of the power supply network corresponding to the original voltage domain after a preset second time period after the switch.

[0085] Specifically, through pre-charging, formal switching, and terminating pre-charging, the impact of power switching on the stability of network interface devices is reduced, preventing logic errors or system resets caused by sudden voltage changes. The first duration can be 100ns. After the first pre-charging duration, the target voltage domain is pre-balanced. The second duration can be 10ns. After the second duration, the power supply network corresponding to the original voltage domain is disconnected, confirming that the power switch of the target voltage domain is fully stable.

[0086] This operation can be implemented by the voltage domain switching controller using the hardware description language Verilog. The relevant code can be expressed as:

[0087] Verilog

[0088] / / voltage domain switching controller

[0089] always @(posedge clk) begin / / Execute when the rising edge of clock clk is triggered

[0090] if (target_domain != current_domain) begin

[0091] enable_precharge <= 1; / / Start precharge phase

[0092] #100ns; / / Charge pre-balance

[0093] switch_control <= target_domain; / / Start formal switching

[0094] #10ns;

[0095] enable_precharge <= 0; / / End precharge phase

[0096] end

[0097] end

[0098] Among them, target_domain refers to the target voltage domain, and current_domain refers to the original voltage domain.

[0099] In one possible design, the multiple power supply networks include: a first power supply network, a second power supply network, and a third power supply network;

[0100] The power supply type of the first power supply network includes: a low voltage dropout linear regulator and a switched capacitor;

[0101] The power supply type of the second power supply network includes: a multi-phase buck converter;

[0102] The power supply type of the third power supply network includes: a direct-connect power management integrated circuit;

[0103] The minimum voltage of the voltage domain corresponding to the second power supply network is greater than the maximum voltage of the voltage domain corresponding to the first power supply network;

[0104] The maximum voltage of the voltage domain corresponding to the second power supply network is smaller than the minimum voltage of the voltage domain corresponding to the third power supply network.

[0105] Specifically, a low-dropout (LDO) regulator is a linear regulated power supply that uses a variable resistor to dissipate excess energy between the input and output voltages, thereby providing a stable, low-noise output voltage. A switched capacitor is an inductorless DC-DC (direct current) converter that uses the charging and discharging of capacitors to transfer energy and achieves step-up, step-down, or inverting voltage conversion by changing the capacitor connection method. A multiphase buck converter (Buck) is a high-performance switch-mode power supply consisting of multiple phases connected in parallel and operating in an interleaved manner. A direct-connect power management integrated circuit (PMIC) is a power management integrated circuit that is mounted next to the network interface device and directly powered through short, wide printed circuit board (PCB) traces or interconnections within the package.

[0106] Table 1 is a parameter table of multiple power supply networks provided in an embodiment of the present application.

[0107] Table 1:

[0108]

[0109] The technical effect of the embodiment of the present application is that power supply requirements under different voltage ranges are met through the first power supply network, the second power supply network and the third power supply network.

[0110] Figure 3 Schematic diagram of the process of the energy efficiency optimization method of the network interface device provided in the embodiment of the present application Figure 3 .like Figure 3 As shown, in one possible design, S109, wherein feedback adjustment of the frequency of the network interface device according to the voltage-frequency adjustment method includes:

[0111] S301: Acquire frequency step information for indicating a frequency adjustment step size.

[0112] If the frequency adjustment step is greater than the preset step threshold, it indicates that a larger range of frequency adjustment is required, and the process continues with S302 ; if the frequency adjustment step is less than or equal to the step threshold, it indicates that fine frequency adjustment is required, and the process continues with S303 .

[0113] S302: Adjust the frequency division coefficient of the pre-configured phase-locked loop module according to the frequency adjustment step size.

[0114] S303: Adjust the delay time of the pre-configured delay locked loop module according to the frequency adjustment step size.

[0115] After executing S302 or S303, continue to execute S304.

[0116] S304: Adjust the frequency of the network interface device through a phase-locked loop module or a delay-locked loop module.

[0117] Specifically, the frequency adjustment step size refers to the specific change or step size (ΔF) required for this frequency adjustment calculated by the server. This value is derived by the energy efficiency optimization algorithm based on factors such as current load, temperature, and predicted energy efficiency ratio.

[0118] The step threshold is a preset frequency change threshold used to distinguish between large and small step adjustments. When the frequency adjustment step size is greater than the step threshold, the frequency is adjusted using a method with lower accuracy but a wider range. Otherwise, the frequency is adjusted using a method with higher accuracy but a narrower range. For example, the step threshold can be 2%.

[0119] A phase-locked loop (PLL) is a circuit used to generate and adjust clock signals. By adjusting its internal frequency division coefficient, the output clock frequency can be changed. PLLs typically offer a wide frequency adjustment range, but the accuracy is relatively low. For example, a PLL can achieve ±15% frequency adjustment with a response time of less than 100ns.

[0120] The division factor is a configurable parameter in a PLL that determines the multiplication or division factor of the input reference clock, directly controlling the output frequency. The division factor is inversely correlated with the frequency adjustment step size. To achieve a larger frequency increase, you typically need to reduce the division factor.

[0121] A Delay-Locked Loop (DLL) is a circuit used to precisely control the phase and delay of a clock signal. By adjusting its internal delay time, the clock period can be fine-tuned. A DLL can achieve high-precision frequency fine-tuning within a small range, but the adjustment range is limited, and the frequency adjustment accuracy of a DLL is lower than that of a PLL. For example, a DLL can compensate for ±2% jitter with an accuracy of 0.1% and a delay of <20ns.

[0122] The frequency division factor is a configurable parameter in the DLL. By increasing or decreasing the delay time, you can fine-tune the clock period. The delay time is positively correlated with the frequency adjustment step size. To increase the frequency, you usually need to decrease the delay time.

[0123] Adjusting the frequency of the network interface device through PLL or DLL can be achieved by the clock adaptive regulator using the hardware description language SystemVerilog. The relevant code can be expressed as:

[0124] SystemVerilog

[0125] module clock_adapter (

[0126] input [3:0] freq_step, / / 0.5% per step, achieving a frequency adjustment step from 0% to 7.5%

[0127] output reg clk_out

[0128] ); / / There are 4 input freq_step and one output clk_out

[0129] always @(*) begin

[0130] if (|freq_step[3:2]) begin / / step size > 2%

[0131] pll_mode <= 1; / / Enable PLL mode

[0132] pll_div <= 64 - freq_step;

[0133] end else begin

[0134] dll_mode <= 1; / / Enable DLL mode

[0135] dll_delay <= freq_step[1:0] * 2;

[0136] end

[0137] end

[0138] endmodule

[0139] Among them, freq_step refers to the frequency adjustment step size, pll_div refers to the frequency division coefficient, and dll_delay refers to the frequency division coefficient.

[0140] The technical effect of the embodiment of the present application is: through a hierarchical adjustment strategy, the wide-range adjustment capability of the phase-locked loop module and the high-precision fine-tuning capability of the delay-locked loop module are combined, so that the server can not only cope with drastic changes in load, but also achieve fine optimization of energy efficiency, taking into account both dynamic response and steady-state accuracy.

[0141] It should be noted that since the voltage adjustment and frequency adjustment of the network interface device are decoupled, the present embodiment does not limit the execution order of S201 to S202 and S301 to S304. S201 to S202 may be executed first, followed by S301 to S304; S301 to S304 may be executed first, followed by S201 to S202; or S201 to S202 and S301 to S304 may be executed simultaneously. Furthermore, when executing S301 to S304, only the PLL mode may be enabled, only the DLL mode may be enabled, or the PLL mode may be enabled first, followed by the DLL mode.

[0142] In one possible design, S102 includes:

[0143] The real-time performance parameters are input into a preconfigured nonlinear multi-order polynomial energy efficiency model, and a predicted energy efficiency ratio of the network interface device is output; wherein the nonlinear multi-order polynomial energy efficiency model is used to describe the nonlinear relationship between the predicted energy efficiency ratio and the real-time performance parameters.

[0144] Furthermore, the real-time performance parameters include: normalized bandwidth utilization U, operating temperature T and average packet length S.

[0145] Normalized bandwidth utilization (U) refers to bandwidth utilization normalized to the interval [0, 1]. Bandwidth utilization is the ratio of the currently used bandwidth to its theoretical maximum bandwidth. When U = 0, the network interface device is completely idle; when U = 1, it is fully loaded.

[0146] Operating temperature T refers to the core temperature of the network interface device during operation, measured in degrees Celsius (°C), which directly affects the physical properties of semiconductor devices.

[0147] The average packet length, S, is the average length of all data packets transmitted by a network interface device over a period of time, measured in bytes. The average packet length is used to enable the nonlinear multi-order polynomial energy efficiency model to perceive and adapt to different network traffic patterns.

[0148] The nonlinear multi-order polynomial energy efficiency model is expressed as:

[0149]

[0150] in, is the predicted energy efficiency ratio of the network interface device; k is an integer between 0 and 3, and k can take the values ​​0, 1, 2, and 3 in sequence; are the preset coefficients respectively; is a preset function of the operating temperature; It is the load change rate obtained based on the normalized bandwidth utilization.

[0151] is a third-order polynomial about the normalized bandwidth utilization, It can be a linear function or a quadratic function, etc., which is used to dynamically reflect the impact of operating temperature on the nonlinear characteristics of energy efficiency.

[0152] It is used to describe the impact of average packet length on energy efficiency. It is a packet length penalty term to match the direct memory access (DMA) engine characteristics of network interface devices. This term shows that as the average packet length increases, the gain in energy efficiency will gradually saturate until it approaches the maximum value. .

[0153] is the absolute value term of the gradient, which is used to suppress false triggering of negative gradients.

[0154] Furthermore, the nonlinear multi-order polynomial energy efficiency model can also be expressed as:

[0155] math

[0156]

[0157] In a possible design, each coefficient is calibrated, and the calibration results are shown in Table 2. Table 2 is a parameter table of each coefficient calibration provided in an embodiment of the present application.

[0158] Table 2:

[0159]

[0160] The technical effect of the embodiment of the present application is: based on the normalized bandwidth utilization, operating temperature and average packet length, a nonlinear multi-order polynomial energy efficiency model is used to determine the predicted energy efficiency ratio of the network interface device, thereby improving the accuracy of the predicted energy efficiency ratio.

[0161] In one possible design, after executing S102, the method further includes:

[0162] S103: Obtain actual energy efficiency ratio.

[0163] If the error between the predicted energy efficiency ratio and the actual energy efficiency ratio is greater than the preset error threshold, it is determined that the nonlinear multi-order polynomial energy efficiency model needs to be calibrated, and S104 is executed; otherwise, it is determined that the nonlinear multi-order polynomial energy efficiency model is sufficiently accurate, and S105 is continued.

[0164] S104, iteratively update according to the error and the preset learning rate , until the error is less than or equal to the error threshold.

[0165] in, .

[0166] Specifically, the error threshold is a pre-set, acceptable maximum error value, which can be set as 3%. The learning rate is used to control the step size or speed of model parameter updates. A larger learning rate makes the model respond more quickly to new data, but may lead to insufficient stability. Conversely, a smaller learning rate makes parameter updates smoother and the convergence process more stable, but slower.

[0167] The server then measures the actual power consumption and performance of the network interface device using hardware sensors or performance counters, and calculates the actual energy efficiency ratio based on this information. The difference between the predicted and actual energy efficiency ratios is then calculated, representing the error between the two. If the nonlinear multi-order polynomial energy efficiency model requires calibration, the model parameters are adjusted based on the current error and the preset learning rate.

[0168] The update rule can be expressed as:

[0169]

[0170] in, The adjustment amount, after adjustment sum. The actual energy efficiency ratio. is the learning rate, The value of can be between 0.00005 and 0.0002, for example, it can be 0.0001; The negative sign before is used to implement negative feedback regulation. Specifically, if , is a negative number, adjusted Will decrease, causing the model's next predicted value to decrease and approach the actual value; conversely, the model's next predicted value will increase.

[0171] Further, The update rule can also be expressed as:

[0172] math

[0173] \Delta a_1 = -0.0001 \times (\eta_{pred} - \eta_{meas})

[0174] The technical effect of the embodiment of the present application is: by introducing negative feedback regulation based on the actual energy efficiency ratio, the long-term effectiveness of the nonlinear multi-order polynomial energy efficiency model after actual deployment is improved.

[0175] In one possible design, the iterative update is performed based on the error and a preset learning rate. and , until the error is less than or equal to the respective error threshold.

[0176] The update method of the above parameters is the same as The updating method is similar and will not be described in detail in the embodiments of this application.

[0177] In a possible design, if the error between the predicted energy efficiency ratio and the actual energy efficiency ratio is greater than a preset error threshold, it is determined that the adjustment of the voltage and frequency of the network interface device does not meet expectations, and a rollback mechanism is executed.

[0178] The rollback mechanism is used to quickly restore the system to a known, stable, and efficient operating state, mitigating risks and ensuring service quality. Specifically, before each voltage / frequency adjustment, the server automatically records the current critical operating state as a baseline. If a rollback is determined, the voltage and frequency of the network interface device are immediately restored to the recorded baseline state before the adjustment. Furthermore, after the rollback is executed, the server remeasures the actual energy efficiency ratio and performance to confirm that a stable state has been restored.

[0179] The technical effect of the embodiment of the present application is: if the error between the predicted energy efficiency ratio and the actual energy efficiency ratio is greater than a preset error threshold, the stability and reliability of the network interface device are improved through the rollback mechanism.

[0180] In one possible design, when performing AVFS adjustment, S108 includes:

[0181] S1081. Obtain an energy efficiency gradient based on the predicted energy efficiency ratio.

[0182] If the energy efficiency gradient is less than the preset first gradient threshold and the load change rate is greater than the preset change rate threshold, execute S1082; if the energy efficiency gradient is greater than the preset second gradient threshold and the operating temperature is greater than the preset temperature threshold, execute S1083.

[0183] S1082: Determine that the voltage and frequency adjustment mode is voltage and frequency boost.

[0184] Among them, the first gradient threshold is a negative number;

[0185] S1083: Determine that the voltage and frequency adjustment mode is voltage and frequency reduction.

[0186] The second gradient threshold is a positive number, and the second gradient threshold is smaller than the absolute value of the first gradient threshold.

[0187] Specifically, the energy efficiency gradient refers to the rate of change of the predicted energy efficiency ratio relative to the normalized bandwidth utilization, that is, the partial derivative , which means that when U changes slightly, The changing trend and rate of , which means that increasing U will improve , the network interface device is in the region where the energy efficiency increases with the load; the negative gradient , which means that increasing U will reduce , the network interface device is in the region where the energy efficiency decreases as the load increases.

[0188] The first gradient threshold is a pre-set, negative energy efficiency gradient threshold, which can be -0.01, and is used to determine whether the network interface device is in a region where energy efficiency decreases as the load increases; the second gradient threshold is a pre-set, positive energy efficiency gradient threshold, which can be 0.005, and is used to determine whether the network interface device is in a region where energy efficiency increases as the load increases; the change rate threshold refers to a pre-set load change rate threshold, which can be 0. When the load change rate is greater than the change rate threshold, it is considered that the load is increasing sharply; the temperature threshold is a pre-set, higher operating temperature threshold, which can be 85°C. When the operating temperature is greater than the temperature threshold, it indicates that the heat dissipation pressure is high, there is a risk of thermal throttling or reliability, and cooling should be given priority.

[0189] The server checks whether the following two conditions are met simultaneously: Condition 1: The energy efficiency gradient is less than the first gradient threshold, and Condition 2: The load change rate is greater than the change rate threshold. If both conditions are met, the voltage and frequency adjustment method is determined to be boost. The goal is to significantly improve performance to break out of the low-efficiency zone and meet sudden performance requirements.

[0190] The server checks whether the following two conditions are simultaneously met: Condition 1: the energy efficiency gradient is greater than the second gradient threshold; Condition 2: the operating temperature is greater than the temperature threshold. If both conditions are met, the voltage and frequency adjustment method is determined to be voltage and frequency reduction. The purpose is to actively reduce temperature by reducing power consumption to prevent overheating. At the same time, because energy efficiency is still increasing, a small frequency reduction will not immediately lead to a significant drop in energy efficiency.

[0191] The above conditional judgment can also be expressed as:

[0192] math

[0193] \text{Decision function} = \begin{cases}

[0194] \text{ } & \text{if} \frac{\partial \eta}{\partial U} < -0.01 \text{ AND} \frac{dU}{dt} > 0 \\

[0195] \text{Reduced voltage and frequency} & \text{if} \frac{\partial \eta}{\partial U} > 0.005\text{ OR} T > 85℃

[0196] \end{cases}

[0197] The technical effect of the embodiment of the present application is: through the energy efficiency gradient, combined with the load change rate and operating temperature, it is determined whether the voltage and frequency adjustment method is to increase the voltage and frequency or to decrease the voltage and frequency.

[0198] In one possible design, the load change rate is calculated as:

[0199]

[0200] in, is the sampling period; is the normalized bandwidth utilization at time t; is the normalized bandwidth utilization three sampling periods before time t.

[0201] In addition, to achieve noise filtering, if , then output the calculated value; otherwise, output 0.

[0202] The calculation formula for load change rate can also be expressed as:

[0203] def efficiency_gradient(U_history):

[0204] grad=(U_history[-1]-U_history[-4]) / (3*Δt)#3 point difference

[0205] return grad if abs(grad) > 0.001 else 0

[0206] In a possible design, when performing dynamic operating point jumping, the method further includes:

[0207] If the normalized bandwidth utilization is within a preset utilization threshold range, and the predicted energy efficiency ratio is less than a preset energy efficiency ratio threshold, it is determined that the network interface device is detected to be in an energy efficiency depression state.

[0208] Accordingly, S106 determines the target voltage based on the real-time performance parameters and the predicted energy efficiency ratio, including:

[0209] If the load change rate is a positive number, execute S1061; if the load change rate is a negative number, execute S1062.

[0210] S1061: Continue to increase the first utilization threshold value for the normalized bandwidth utilization until the modified normalized bandwidth utilization is outside the utilization threshold range.

[0211] S1062: Continue to subtract the second utilization threshold from the normalized bandwidth utilization until the modified normalized bandwidth utilization is outside the utilization threshold range.

[0212] After executing S1061 or S1062, continue to execute S1063.

[0213] S1063 : Obtain a target voltage according to the modified normalized bandwidth utilization and the predicted energy efficiency ratio.

[0214] Specifically, the utilization threshold interval is a pre-set range of utilization values, such as [20%, 70%]. This range is identified as an area where network interface devices are prone to falling into energy efficiency troughs. The first utilization threshold and the second utilization threshold represent two pre-set utilization increments, respectively. The first utilization threshold can be set to 15%, and the second utilization threshold can be set to 10%. Both thresholds serve as step adjustments to escape energy efficiency troughs. Specifically, the first utilization threshold is used to increase the load, while the second utilization threshold is used to reduce the load.

[0215] If the load change rate is positive, the server considers that the load is on the rise and adopts an upward escape strategy to escape the energy efficiency trough. The current U value is continuously increased by the first utilization threshold until the modified U value exceeds the maximum value of the utilization threshold interval.

[0216] If the load change rate is negative, the server believes that the load has a trend of continuing to decrease, and then the second utilization threshold is continuously subtracted from the current U value until the modified U is less than the minimum value of the utilization threshold interval.

[0217] The determination of energy efficiency low state can be implemented in C language. The relevant code can be expressed as follows:

[0218] c

[0219] if (U_current ∈ [20%,70%] && η_current < η_threshold) {

[0220] target_U = (dU_dt > 0) ? U_current + 15% : U_current - 10%;

[0221] enforce_voltage_freq(target_U);

[0222] }

[0223] After obtaining the modified normalized bandwidth utilization, the target voltage required to support the modified normalized bandwidth utilization is reversely derived through model calculation or table lookup.

[0224] Furthermore, in the timing control for adjusting the voltage of the network interface device according to the target voltage, the delay of voltage domain switching is 18µs, the delay of frequency locking is 2µs, and the delay of stability detection is 5µs.

[0225] The technical effect of the embodiments of the present application is: by intelligently selecting the escape direction and calculating the target voltage, the network interface device is removed from the energy efficiency trough state, thereby improving the overall energy efficiency performance of the network interface device.

[0226] In one possible design, after determining the target voltage according to the real-time performance parameter and the predicted energy efficiency ratio in S106, the method further includes:

[0227] S1064: Obtain a compensation voltage according to the operating temperature and the preset nominal voltage.

[0228] S1065: Correct the target voltage according to the compensation voltage.

[0229] Specifically, the nominal voltage is the rated operating voltage at a reference temperature; for example, a reference temperature of 25°C corresponds to a nominal voltage of 0.9V. The compensation voltage is positively correlated with the difference between the operating temperature and the reference temperature. If the operating temperature is higher than the reference temperature, the compensation voltage is typically positive, indicating that the voltage needs to be increased to compensate for the degradation of transistor performance at high temperatures and maintain timing correctness. If the operating temperature is lower than the reference temperature, the compensation voltage is typically negative or zero, indicating that the voltage can be reduced because transistor performance is better at lower temperatures.

[0230] Compensation voltage The calculation formula can be expressed as:

[0231]

[0232] in, is the nominal voltage, is the reference temperature; is the preset compensation coefficient, which indicates the voltage adjustment required for every 1°C change in temperature. The value can be 0.8mV / ℃.

[0233] Compensation voltage The calculation formula can also be expressed as:

[0234] math

[0235] V_{adj} = V_{nom} + (T - 25℃) \times 0.0008 \text{ V / ℃}

[0236] The technical effect of the embodiment of the present application is: by introducing a temperature compensation voltage to correct the preliminary target voltage, the practicality and robustness of the network interface device are improved.

[0237] In a possible design, when performing dynamic operating point jumping, the method further includes:

[0238] A frequency threshold is obtained according to the voltage of the network interface device.

[0239] The frequency threshold is positively correlated with the voltage of the network interface device.

[0240] If the frequency of the network interface device is greater than the frequency threshold, the preset safety protection mechanism is triggered.

[0241] Specifically, the frequency threshold is a maximum clock frequency upper limit at which the network interface device can operate safely and stably, which is dynamically calculated based on the current power supply voltage.

[0242] The server uses a hardware watchdog to check for illegal voltage and frequency combinations every 1µs. Based on the voltage of the network interface device, it uses a preset voltage-frequency mapping relationship, lookup table, or mathematical function to dynamically calculate the corresponding safe frequency threshold at the current voltage. This mapping relationship is usually derived from the process, voltage, and temperature characteristic testing of the network interface device to ensure that all timing paths can meet the setup / hold time requirements at a given voltage.

[0243] If the frequency of the network interface device exceeds the frequency threshold, the preset safety protection mechanism is triggered. The safety protection mechanism is a set of pre-set emergency measures to prevent the network interface device from being damaged due to overload or instability. The safety protection mechanism may include:

[0244] Forced frequency reduction means immediately reducing the operating frequency to a safe level.

[0245] Voltage boosting means temporarily increasing the supply voltage to match the current high frequency demand if the power supply allows.

[0246] Alarm and logging, which means sending alarms to system management software and recording event logs;

[0247] Traffic throttling, which temporarily reduces the data processing rate to reduce the load on the device;

[0248] System reset, which restarts the device as a last resort in extremely unstable situations.

[0249] The technical effect of the embodiment of the present application is: by establishing a positive correlation between voltage and frequency thresholds, whether the frequency exceeds the limit is monitored in real time, thereby preventing system instability caused by insufficient voltage.

[0250] In one possible design, the frequency threshold The calculation formula can be expressed as:

[0251]

[0252] Where V is the voltage of the network interface device.

[0253] The calculation formula for the frequency threshold can also be expressed as:

[0254] math

[0255] f_{max}(V) = \begin{cases}

[0256] 0.8 \text{GHz} & \text{if} V < 0.7v \\

[0257] 1.5 \text{GHz} & \text{if} V \geq 1.0v

[0258] \end{cases}

[0259] Table 3 is a parameter table for the energy efficiency optimization method and the autoregressive integral moving average model algorithm for network interface devices provided in the embodiments of the present application. As shown in Table 3, the parameters of the two methods in terms of dealing with burst traffic, energy efficiency stability, controlling temperature drift, and controlling hardware resource usage are shown. It can be seen that the energy efficiency optimization method for network interface devices has a significant improvement over the autoregressive integral moving average model algorithm.

[0260] Table 3:

[0261]

[0262] Figure 4 This is a schematic diagram of the structure of the energy efficiency optimization device of the network interface device provided in the embodiment of the present application. Figure 4 As shown, in an embodiment of the present application, the energy efficiency optimization device of the network interface device may be located in an electronic device, and the device includes:

[0263] A first acquisition module 410 is used to acquire real-time performance parameters of a network interface device;

[0264] an energy efficiency ratio calculation module 420 for determining a predicted energy efficiency ratio of the network interface device based on real-time performance parameters of the network interface device;

[0265] A depression detection module 430 is used to detect whether the network interface device is in an energy efficiency depression state based on real-time performance parameters and predicted energy efficiency ratio;

[0266] The target value calculation module 440 is configured to determine a target voltage and a target frequency based on the real-time performance parameters and the predicted energy efficiency ratio when the network interface device is detected to be in an energy efficiency trough state. The target voltage and target frequency are the voltage and frequency when the network interface device exits the energy efficiency trough state.

[0267] a jump adjustment module 450 for adjusting the voltage and frequency of the network interface device according to the target voltage and target frequency using a dynamic operating point jump mechanism;

[0268] An adjustment determination module 460 is configured to determine a voltage and frequency adjustment method based on real-time performance parameters and a predicted energy efficiency ratio if it is detected that the network interface device is not in an energy efficiency trough state;

[0269] The adaptive adjustment module 470 is used to adjust the voltage and frequency of the network interface device according to the voltage and frequency adjustment mode if the voltage and frequency adjustment mode is determined to be voltage and frequency increase or voltage and frequency decrease, until the voltage and frequency adjustment mode is determined to be stop adjustment.

[0270] The energy efficiency optimization device of the network interface device provided in the embodiment of the present application can perform Figure 1 The technical solution of the method embodiment shown in the figure has the same implementation principle and technical effect as Figure 1 The method embodiments shown are similar and will not be described in detail in the embodiments of this application.

[0271] At the same time, the energy efficiency optimization device for a network interface device provided in an embodiment of the present application is further refined based on the energy efficiency optimization device for a network interface device provided in an embodiment of the previous application.

[0272] In one possible design, the adaptive adjustment module 470 includes:

[0273] A voltage domain selection module is used to determine an adjusted target voltage domain from a plurality of preset voltage domains according to the voltage-frequency adjustment method; wherein each voltage domain corresponds to an independent power supply network;

[0274] The voltage domain switching module is used to switch to the power supply network corresponding to the target voltage domain for power supply if it is detected that the current original voltage domain is inconsistent with the target voltage domain.

[0275] In a possible design, the voltage domain switching module includes:

[0276] A pre-charging module, configured to pre-charge the power supply network corresponding to the target voltage domain until a preset first time duration is reached;

[0277] The network switching module is used to switch to the power supply network corresponding to the target voltage domain for power supply, and disconnect the power supply connection of the power supply network corresponding to the original voltage domain after a preset second time period after the switch.

[0278] In one possible design, the multiple power supply networks include: a first power supply network, a second power supply network, and a third power supply network;

[0279] The power supply type of the first power supply network includes: a low voltage dropout linear regulator and a switched capacitor;

[0280] The power supply type of the second power supply network includes: a multi-phase buck converter;

[0281] The power supply type of the third power supply network includes: a direct-connect power management integrated circuit;

[0282] The minimum voltage of the voltage domain corresponding to the second power supply network is greater than the maximum voltage of the voltage domain corresponding to the first power supply network;

[0283] The maximum voltage of the voltage domain corresponding to the second power supply network is smaller than the minimum voltage of the voltage domain corresponding to the third power supply network.

[0284] In one possible design, the adaptive adjustment module 470 includes:

[0285] a step determination module, configured to obtain frequency step information indicating a frequency adjustment step size;

[0286] A phase-locked loop adjustment module, configured to adjust a frequency division coefficient of a preconfigured phase-locked loop module according to the frequency adjustment step size if the frequency adjustment step size is greater than a preset step size threshold; wherein the frequency division coefficient is negatively correlated with the frequency adjustment step size;

[0287] a delay-locked loop adjustment module configured to adjust a preconfigured delay time of the delay-locked loop module according to the frequency adjustment step size if the frequency adjustment step size is less than or equal to the step size threshold; wherein the delay time is positively correlated with the frequency adjustment step size; and the frequency adjustment accuracy of the phase-locked loop module is lower than that of the delay-locked loop module;

[0288] The frequency adjustment module is used to adjust the frequency of the network interface device through the phase-locked loop module or the delay-locked loop module.

[0289] In one possible design, the energy efficiency ratio calculation module 420 is used to input the real-time performance parameters into a preconfigured nonlinear multi-order polynomial energy efficiency model and output the predicted energy efficiency ratio of the network interface device; wherein the nonlinear multi-order polynomial energy efficiency model is used to describe the nonlinear relationship between the predicted energy efficiency ratio and the real-time performance parameters.

[0290] In one possible design, the real-time performance parameters include: normalized bandwidth utilization U, operating temperature T, and average packet length S;

[0291] The nonlinear multi-order polynomial energy efficiency model is expressed as:

[0292]

[0293] in, Predicted energy efficiency ratio for network interface devices; are the preset coefficients respectively; is a preset function of the operating temperature; It is the load change rate obtained based on the normalized bandwidth utilization.

[0294] In one possible design, it also includes:

[0295] The second acquisition module is used to obtain the actual energy efficiency ratio;

[0296] The parameter update module is used to iteratively update the parameter according to the error and the preset learning rate if the error between the predicted energy efficiency ratio and the actual energy efficiency ratio is greater than the preset error threshold. , until the error is less than or equal to the error threshold.

[0297] In one possible design, the adjustment determination module 460 includes:

[0298] A gradient calculation module is used to obtain an energy efficiency gradient based on the predicted energy efficiency ratio;

[0299] A first determining module is configured to determine that the voltage-frequency adjustment mode is voltage-and-frequency boosting if the energy efficiency gradient is less than a preset first gradient threshold and the load change rate is greater than a preset change rate threshold; wherein the first gradient threshold is a negative number;

[0300] The second determination module is used to determine that the voltage and frequency adjustment method is voltage and frequency reduction if the energy efficiency gradient is greater than a preset second gradient threshold and the operating temperature is greater than a preset temperature threshold; wherein the second gradient threshold is a positive number and the second gradient threshold is less than the absolute value of the first gradient threshold.

[0301] In one possible design, it also includes:

[0302] a third determining module, configured to determine that the network interface device is in an energy efficiency depression state if the normalized bandwidth utilization is within a preset utilization threshold range and the predicted energy efficiency ratio is less than a preset energy efficiency ratio threshold;

[0303] Accordingly, the target value calculation module 440 includes:

[0304] a first modifying module, configured to continuously increase the normalized bandwidth utilization by a first utilization threshold if the load change rate is a positive number, until the modified normalized bandwidth utilization is outside the utilization threshold interval;

[0305] a second modifying module, configured to, if the load change rate is a negative number, continue to subtract a second utilization threshold from the normalized bandwidth utilization until the modified normalized bandwidth utilization is outside the utilization threshold range;

[0306] The voltage calculation module is used to obtain a target voltage according to the modified normalized bandwidth utilization and the predicted energy efficiency ratio.

[0307] In one possible design, it also includes:

[0308] The compensation calculation module is used to obtain a compensation voltage based on the operating temperature and a preset nominal voltage; wherein the nominal voltage is the rated operating voltage at the reference temperature; and the compensation voltage is positively correlated with the difference between the operating temperature and the reference temperature;

[0309] The voltage repair module is used to correct the target voltage according to the compensation voltage.

[0310] In one possible design, it also includes:

[0311] A threshold calculation module, configured to obtain a frequency threshold according to a voltage of the network interface device; wherein the frequency threshold is positively correlated with the voltage of the network interface device;

[0312] The security protection module is used to trigger a preset security protection mechanism if the frequency of the network interface device is greater than a frequency threshold.

[0313] The energy efficiency optimization device of the network interface device provided in the embodiment of the present application can perform Figure 2 and Figure 3 The technical solution of the method embodiment shown in the figure has the same implementation principle and technical effect as Figure 2 and Figure 3 The method embodiments shown are similar and will not be described in detail in the embodiments of this application.

[0314] Figure 5 This is a schematic diagram of the structure of the electronic device provided in this application. Figure 5 As shown, the electronic device provided by this embodiment includes: at least one processor 510 and a memory 520. Optionally, the electronic device further includes a communication component 530. The processor 510, the memory 520 and the communication component 530 are connected via a bus.

[0315] During the specific implementation process, at least one processor 510 executes the computer-executable instructions stored in the memory 520, so that the at least one processor 510 executes the above-mentioned embodiment of the method for optimizing the energy efficiency of the network interface device.

[0316] The specific implementation process of the processor 510 can be found in the above method embodiment. Its implementation principle and technical effects are similar and will not be repeated here in this embodiment.

[0317] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the application may be directly executed by a hardware processor or by a combination of hardware and software modules within the processor.

[0318] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage.

[0319] A bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be categorized as address buses, data buses, and control buses. For ease of illustration, the buses in the drawings of this application are not limited to just one bus or just one type of bus.

[0320] An embodiment of the present application further provides a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the steps of any of the above-mentioned embodiments of the method for optimizing energy efficiency of a network interface device when running.

[0321] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0322] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any of the above-mentioned embodiments of the method for optimizing energy efficiency of a network interface device are implemented.

[0323] An embodiment of the present application also provides another computer program product, including a non-volatile computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps in any of the above-mentioned energy efficiency optimization method embodiments of the network interface device.

[0324] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0325] The above is a detailed introduction to the energy efficiency optimization method, device, medium and product of a network interface device provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only applicable to help understand the method and core idea of ​​the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

Claims

1. A method for optimizing energy efficiency of a network interface device, characterized in that: include: Obtain real-time performance parameters of network interface devices; determining a predicted energy efficiency ratio of the network interface device based on real-time performance parameters of the network interface device; detecting, based on the real-time performance parameter and the predicted energy efficiency ratio, whether the network interface device is in an energy efficiency trough state; If it is detected that the network interface device is in the energy efficiency trough state, determining a target voltage and a target frequency according to the real-time performance parameter and the predicted energy efficiency ratio; wherein the target voltage and the target frequency are the voltage and frequency when the network interface device exits the energy efficiency trough state; Adopting a dynamic operating point jump mechanism to adjust the voltage and frequency of the network interface device according to the target voltage and the target frequency; If it is detected that the network interface device is not in the energy efficiency trough state, determining a voltage and frequency adjustment method according to the real-time performance parameter and the predicted energy efficiency ratio; If it is determined that the voltage and frequency adjustment mode is voltage and frequency increase or voltage and frequency reduction, the voltage and frequency of the network interface device are adjusted according to the voltage and frequency adjustment mode until it is determined that the voltage and frequency adjustment mode is to stop adjustment.

2. The method according to claim 1, characterized in that The step of feedback-adjusting the voltage of the network interface device according to the voltage-frequency adjustment method includes: According to the voltage-frequency adjustment method, determining an adjusted target voltage domain from a plurality of preset voltage domains; wherein each of the voltage domains corresponds to an independent power supply network; If it is detected that the current original voltage domain is inconsistent with the target voltage domain, the power supply is switched to the power supply network corresponding to the target voltage domain for power supply.

3. The method according to claim 2, characterized in that The switching to the power supply network corresponding to the target voltage domain for power supply includes: Pre-charging the power supply network corresponding to the target voltage domain until a preset first time duration is reached; Switch to the power supply network corresponding to the target voltage domain for power supply, and after a preset second time period after the switch, disconnect the power supply connection of the power supply network corresponding to the original voltage domain.

4. The method according to claim 3, characterized in that The plurality of power supply networks include: a first power supply network, a second power supply network and a third power supply network; The power supply type of the first power supply network includes: a low voltage dropout linear regulator and a switched capacitor; The power supply type of the second power supply network includes: a multi-phase buck converter; The power supply type of the third power supply network includes: a direct-connect power management integrated circuit; The minimum voltage of the voltage domain corresponding to the second power supply network is greater than the maximum voltage of the voltage domain corresponding to the first power supply network; A maximum voltage of a voltage domain corresponding to the second power supply network is smaller than a minimum voltage of a voltage domain corresponding to the third power supply network.

5. The method according to claim 1, wherein The step of feedback-adjusting the frequency of the network interface device according to the voltage-frequency adjustment method includes: Obtaining frequency step information for indicating a frequency adjustment step size; If the frequency adjustment step is greater than a preset step threshold, adjusting the frequency division coefficient of the preconfigured phase-locked loop module according to the frequency adjustment step; wherein the frequency division coefficient is negatively correlated with the frequency adjustment step; If the frequency adjustment step is less than or equal to the step threshold, adjusting the delay time of the preconfigured delay locked loop module according to the frequency adjustment step; wherein the delay time is positively correlated with the frequency adjustment step; and the frequency adjustment accuracy of the phase locked loop module is lower than the frequency adjustment accuracy of the delay locked loop module; The frequency of the network interface device is adjusted by the phase-locked loop module or the delay-locked loop module.

6. The method according to claim 1, characterized in that Determining the predicted energy efficiency ratio of the network interface device according to the real-time performance parameters of the network interface device includes: The real-time performance parameters are input into a preconfigured nonlinear multi-order polynomial energy efficiency model, and a predicted energy efficiency ratio of the network interface device is output; wherein the nonlinear multi-order polynomial energy efficiency model is used to describe the nonlinear relationship between the predicted energy efficiency ratio and the real-time performance parameters.

7. The method according to claim 6, characterized in that The real-time performance parameters include: normalized bandwidth utilization U, operating temperature T, and average packet length S; The nonlinear multi-order polynomial energy efficiency model is expressed as: in, a predicted energy efficiency ratio for the network interface device; are the preset coefficients respectively; is a preset function related to the working temperature; is the load change rate obtained according to the normalized bandwidth utilization.

8. The method according to claim 7, characterized in that After determining the predicted energy efficiency ratio of the network interface device according to the real-time performance parameters of the network interface device, the method further includes: Obtain actual energy efficiency ratio; If the error between the predicted energy efficiency ratio and the actual energy efficiency ratio is greater than a preset error threshold, the iterative update is performed based on the error and the preset learning rate. , until the error is less than or equal to the error threshold.

9. The method according to claim 7, characterized in that The determining of the voltage and frequency adjustment method according to the real-time performance parameter and the predicted energy efficiency ratio includes: Obtaining an energy efficiency gradient according to the predicted energy efficiency ratio; If the energy efficiency gradient is less than a preset first gradient threshold, and the load change rate is greater than a preset change rate threshold, then the voltage-frequency adjustment mode is determined to be voltage- and frequency-boosting; wherein the first gradient threshold is a negative number; If the energy efficiency gradient is greater than a preset second gradient threshold and the operating temperature is greater than a preset temperature threshold, the voltage and frequency adjustment method is determined to be voltage and frequency reduction; wherein the second gradient threshold is a positive number and the second gradient threshold is less than the absolute value of the first gradient threshold.

10. The method according to claim 7, characterized in that Also includes: If the normalized bandwidth utilization is within a preset utilization threshold range, and the predicted energy efficiency ratio is less than a preset energy efficiency ratio threshold, it is determined that the network interface device is detected to be in the energy efficiency trough state; Accordingly, determining the target voltage according to the real-time performance parameter and the predicted energy efficiency ratio includes: If the load change rate is a positive number, continuously increasing the first utilization threshold value to the normalized bandwidth utilization until the modified normalized bandwidth utilization is outside the utilization threshold range; If the load change rate is a negative number, continuously subtracting the second utilization threshold from the normalized bandwidth utilization until the modified normalized bandwidth utilization is outside the utilization threshold range; A target voltage is obtained according to the modified normalized bandwidth utilization and the predicted energy efficiency ratio.

11. The method according to claim 10, characterized in that After determining the target voltage according to the real-time performance parameter and the predicted energy efficiency ratio, the method further includes: Obtaining a compensation voltage according to the operating temperature and a preset nominal voltage; wherein the nominal voltage is a rated operating voltage at a reference temperature; and the compensation voltage is positively correlated with a difference between the operating temperature and the reference temperature; The target voltage is corrected according to the compensation voltage.

12. The method according to claim 10, characterized in that Also includes: Obtaining a frequency threshold according to the voltage of the network interface device; wherein the frequency threshold is positively correlated with the voltage of the network interface device; If the frequency of the network interface device is greater than the frequency threshold, a preset safety protection mechanism is triggered.

13. An electronic device, characterized in that: include: Memory for storing computer programs; A processor is configured to implement the steps of the method for optimizing the energy efficiency of a network interface device according to any one of claims 1 to 12 when executing the computer program.

14. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of the method for optimizing energy efficiency of a network interface device according to any one of claims 1 to 12.

15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for optimizing energy efficiency of a network interface device according to any one of claims 1 to 12 are implemented.

Images