Semiconductor device power consumption simulation data analysis method and system

Abstract

The application relates to the technical field of semiconductor simulation, in particular to a semiconductor device power consumption simulation data analysis method and system, which comprises the following steps: setting a plurality of working scenes and design parameters of a semiconductor device, simulating the semiconductor device under each working scene and each design parameter, obtaining a temperature sensitivity coefficient based on the difference between the junction temperature data and the environmental temperature data of the semiconductor device under any working scene and in combination with the dynamic proportion under the working scene, adjusting the static power consumption data change condition of the semiconductor device under different design parameters, and respectively performing weighted processing on the dynamic proportion adjustment result under the working scene and the change condition of the dynamic power consumption data, so that the difference between the weighted static power consumption change amount and the weighted dynamic power consumption change amount is used to iteratively optimize the design parameters of the semiconductor device. The application effectively reduces the overall power consumption of the semiconductor device and enhances the thermal stability.

Description

Technical Field

[0001] This invention relates to the field of semiconductor simulation technology, and more specifically to a method and system for analyzing power consumption simulation data of semiconductor devices. Background Technology

[0002] In semiconductor device design, power consumption simulation is an indispensable core step that directly affects the feasibility of the design and the competitiveness of the final product. As processes iterate towards smaller sizes and higher integration, the trade-off between device power consumption, performance, and reliability becomes increasingly prominent, making accurate power consumption simulation a key prerequisite for design optimization.

[0003] Traditional semiconductor device power consumption data analysis revolves around separate static and dynamic power consumption. Current semiconductor device power consumption simulation is mainly divided into two categories, corresponding to power consumption characteristics under different operating conditions: one is static power consumption simulation, focusing on static operating scenarios dominated by leakage current, calculated through DC simulation; the other is dynamic power consumption simulation, targeting transient operating scenarios dominated by capacitor charging / discharging and short-circuit current, acquiring data through transient simulation and analyzing it using specialized tools. These traditional analysis methods have significant limitations in the parameter optimization and iteration process of overall device design, deviating considerably from real-world application scenarios. Existing solutions often treat the two types of simulations separately, relying on fixed process parameters and a single operating condition for optimization, failing to fully consider the coupling and interaction between alternating operating conditions, temperature fluctuations, and design parameters in real-world scenarios, and ignoring the dynamic characteristics changes during device operation. This deviation leads to distorted power consumption analysis results, failing to accurately reflect the actual operating state of the device, thus misleading the iteration direction of design parameters and resulting in optimization solutions lacking comprehensiveness and specificity. Summary of the Invention

[0004] This invention provides a method and system for analyzing power consumption simulation data of semiconductor devices to solve existing problems.

[0005] The semiconductor device power consumption simulation data analysis method and system of the present invention adopts the following technical solution:

[0006] One embodiment of the present invention provides a method for analyzing power consumption simulation data of semiconductor devices, the method comprising the following steps:

[0007] Several operating scenarios and design parameters of semiconductor devices are set, and the semiconductor devices under each operating scenario and design parameter are simulated to obtain the dynamic proportion of the operating scenario, junction temperature data, ambient temperature data, static power consumption data, and dynamic power consumption data of the semiconductor devices during the simulation process.

[0008] Based on the difference between junction temperature data and ambient temperature data of semiconductor devices under any working scenario, and combined with the dynamic proportion under the working scenario, the temperature sensitivity coefficient of semiconductor devices under the working scenario is obtained.

[0009] The static power consumption data of semiconductor devices under different design parameters is adjusted by using a temperature sensitivity coefficient. The adjustment results of the static power consumption data and the dynamic power consumption data under the working scenario are weighted and processed to obtain the weighted static power consumption change and weighted dynamic power consumption change of semiconductor devices under each design parameter in the working scenario.

[0010] The design parameters of the semiconductor device are iteratively optimized based on the difference between the weighted static power consumption change and the weighted dynamic power consumption change to obtain the optimal design parameters of the semiconductor device under the stated operating scenario.

[0011] Optionally, the specific method for obtaining the temperature sensitivity coefficient is as follows:

[0012] Any working scenario is recorded as the target working scenario. The dynamic proportion of the target working scenario is obtained. The dynamic temperature difference coefficient of the semiconductor device in the target working scenario is calculated by combining the difference between the junction temperature data of the semiconductor device and the ambient temperature data in the target working scenario.

[0013] Based on the dynamic temperature difference coefficient of the semiconductor device in the target working scenario and the ambient temperature coefficient in the target working scenario, a reference working scenario is selected from all working scenarios.

[0014] Acquire the static power consumption data of the semiconductor device under the target operating scenario, and record the static power consumption data of the semiconductor device under the reference operating scenario as the reference static power consumption data of the semiconductor device. Based on the difference between the static power consumption data of the semiconductor device under the target operating scenario and the reference static power consumption data, and combined with the dynamic temperature difference coefficient of the semiconductor device under the target operating scenario, calculate the temperature sensitivity coefficient of the semiconductor device under the target operating scenario.

[0015] Optionally, the specific method for obtaining the dynamic temperature difference coefficient is as follows:

[0016] The difference between the junction temperature data of the semiconductor device and the ambient temperature under the target working scenario is denoted as the temperature difference factor of the semiconductor device. The temperature difference factor is corrected by the dynamic proportion under the target working scenario to obtain the dynamic temperature difference coefficient of the semiconductor device under the target working scenario. The dynamic proportion and the temperature difference factor are both positively correlated with the dynamic temperature difference coefficient.

[0017] Optionally, the specific method for obtaining the reference work scenario is as follows:

[0018] Any work scenario includes several scenario parameters;

[0019] By combining the ambient temperature data under the target working scenario and the dynamic temperature difference coefficient of the semiconductor device under the target working scenario, the comprehensive temperature of the semiconductor device under the target working scenario is obtained. The ambient temperature data with the smallest absolute value of the difference between the ambient temperature data corresponding to all working scenarios and the comprehensive temperature is obtained and recorded as the reference ambient temperature. The working scenario to which the reference ambient temperature belongs is obtained, and the working scenario is consistent with all other scenario parameters in the target working scenario except for the ambient temperature data. The working scenario is recorded as the reference working scenario.

[0020] Optionally, the specific method for obtaining the weighted static power consumption change and the weighted dynamic power consumption change is as follows:

[0021] Data interpolation is performed on the static power consumption data and dynamic power consumption data corresponding to all design parameters of the semiconductor device under the target operating environment to obtain the second design-static power consumption sequence and the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment.

[0022] Based on the differences in static power consumption data and dynamic power consumption data between any design parameter and adjacent design parameters in the second design-static power consumption sequence and the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment, determine the static power consumption change and dynamic power consumption change of the corresponding design parameter of the semiconductor device under the target operating environment.

[0023] By utilizing the differences between junction temperature data and ambient temperature data, as well as the temperature sensitivity coefficient, the static power consumption data variation of semiconductor devices under arbitrary design parameters is adjusted and corrected to obtain the corrected static power consumption variation. The dynamic proportion under the target operating environment is used as a weight to weight the dynamic power consumption variation and the corrected static power consumption variation respectively, so as to obtain the weighted static power consumption variation and weighted dynamic power consumption variation of semiconductor devices under the target operating scenario and with the target design parameters as the design parameters.

[0024] Optionally, the specific methods for obtaining the second design-static power consumption sequence and the second design-dynamic power consumption sequence are as follows:

[0025] A sequence of static power consumption data corresponding to all design parameters of a semiconductor device under a target operating environment is obtained and denoted as the first design-static power consumption sequence of the semiconductor device under the target operating environment. The first design-static power consumption sequence is interpolated using the least squares method to obtain the second design-static power consumption sequence of the semiconductor device under the target operating environment. The static power consumption data in the second design-static power consumption sequence is replaced with dynamic power consumption data to obtain the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment.

[0026] Optionally, the specific method for obtaining the static power consumption change and the dynamic power consumption change is as follows:

[0027] Will As a semiconductor device, in the second design-static power consumption sequence, The static power consumption variation of each design parameter, where This indicates the semiconductor device in the second design-static power consumption sequence. Static power consumption data for each design parameter. This indicates the semiconductor device in the second design-static power consumption sequence. The static power consumption data of each design parameter; the second design-static power consumption sequence in the method for obtaining the static power consumption change is replaced with the second design-dynamic power consumption sequence, thereby obtaining the semiconductor device in the second design-dynamic power consumption sequence. The dynamic power consumption change of each design parameter.

[0028] Optionally, the specific method for obtaining the corrected static power consumption change is as follows:

[0029] Design parameters of arbitrary size are denoted as target design parameters. The difference between the junction temperature data of the semiconductor device and the ambient temperature data corresponding to the target operating environment is obtained when the design parameters are the target design parameters under the target operating environment. This difference is denoted as the temperature variable factor of the semiconductor device under the target operating environment and with the target design parameters as the design parameters. A correction factor is obtained by combining the temperature variable factor and the temperature sensitivity coefficient. The correction factor is used to correct the static power consumption change of the semiconductor device under the target design parameters and with the target design parameters as the design parameters. The corrected static power consumption change of the semiconductor device under the target operating environment and with the target design parameters as the design parameters is obtained. The correction factor is negatively correlated with the corrected static power consumption change, and the static power consumption change is positively correlated with the corrected static power consumption change.

[0030] Optionally, the specific method for obtaining the optimal design parameters is as follows:

[0031] Based on the difference between the weighted static power consumption change and the weighted dynamic power consumption change of the semiconductor device under the target operating scenario and with the target design parameters as the design parameters, calculate the power consumption balance coefficient of the semiconductor device under the target operating scenario and with the target design parameters as the design parameters.

[0032] The design parameters of the semiconductor device under the target operating scenario are iteratively updated to obtain the power consumption balance coefficient of each design parameter of the semiconductor device under the target operating scenario; a qualified range is preset, and when the value of the power consumption balance coefficient is within the qualified range, the design parameter corresponding to the power consumption balance coefficient is taken as the optimal design parameter of the semiconductor device under the target operating scenario.

[0033] A semiconductor device power consumption simulation data analysis system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of any one of the semiconductor device power consumption simulation data analysis methods.

[0034] The beneficial effects of the technical solution of this invention are as follows: By constructing a temperature sensitivity coefficient that integrates junction temperature-ambient temperature difference and the dynamic proportion of the working scenario, the nonlinear influence of temperature on static power consumption is quantified; then, using the dynamic proportion as the weight, the temperature-corrected static power consumption change and the dynamic power consumption change are weighted separately to comprehensively characterize the multi-scenario power consumption characteristics of semiconductor devices in real-world usage environments; finally, the design parameters are iteratively optimized based on the differences in the weighted power consumption changes, thereby significantly improving the power consumption simulation and design parameter optimization effects for semiconductor devices. Specifically, the temperature sensitivity coefficient eliminates the bias of the isothermal assumption, making the static power consumption analysis closely match the actual thermal conditions; the dynamic proportion weighting mechanism incorporates the frequency of scenario usage into the optimization objective, ensuring that the design parameters are optimally energy-efficient in typical applications, avoiding "over-design" or insufficient scenario adaptability; the weighted difference iteration achieves a synergistic balance between static leakage current and dynamic switching power consumption, effectively reducing the overall power consumption of semiconductor devices and enhancing thermal stability and long-term reliability. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a flowchart of the steps in the semiconductor device power consumption simulation data analysis method of the present invention;

[0037] Figure 2 This is a block diagram of the semiconductor device power consumption simulation data analysis system of the present invention. Detailed Implementation

[0038] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the semiconductor device power consumption simulation data analysis method and system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0039] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0040] The following description, in conjunction with the accompanying drawings, details the specific scheme of the semiconductor device power consumption simulation data analysis method and system provided by this invention.

[0041] Please see Figure 1 The diagram illustrates a flowchart of a semiconductor device power consumption simulation data analysis method according to an embodiment of the present invention. The method includes the following steps:

[0042] Step S001: Set several operating scenarios and design parameters for the semiconductor device, simulate the semiconductor device under each operating scenario and design parameter, and obtain the dynamic proportion of the operating scenario, junction temperature data, ambient temperature data, static power consumption data, and dynamic power consumption data of the semiconductor device during the simulation process.

[0043] It should be noted that semiconductor device power consumption simulation mainly includes four stages: defining operating conditions and process parameters, clarifying the device's operating scenario, load conditions, temperature range, and core process parameters; segmented power consumption simulation, using specialized tools to simulate static and dynamic power consumption performance separately; data acquisition and analysis, uncovering power distribution characteristics and key constraints; and iterative optimization of design parameters, adjusting the design scheme based on simulation results, and repeating simulation verification until the requirements are met. When analyzing semiconductor device power consumption simulation data, to simulate the real power consumption situation as accurately as possible using power consumption simulation data under two different conditions, there is an inherent connection with the actual usage scenario and conditions of the semiconductor device. Therefore, it is necessary to first obtain simulation results under different modes separately, and at the same time, preset scenario parameters according to the scenario to simulate typical real usage conditions as accurately as possible. Therefore, in order to effectively analyze the power consumption of semiconductor devices, this embodiment of the invention first sets the operating scenario of the semiconductor device and obtains the operating data of the semiconductor device under each operating scenario.

[0044] Specifically, in order to implement the semiconductor device power consumption simulation data analysis method proposed in this embodiment, it is first necessary to set several operating scenarios and design parameters for the semiconductor device, and then simulate the semiconductor device under each operating scenario and design parameter. The specific process is as follows:

[0045] First, several operating scenarios for semiconductor devices are defined. Each operating scenario includes several scenario parameters, which at least include: the operating frequency of the semiconductor device, ambient temperature data, the power supply voltage of the semiconductor device, and dynamic percentage. The dynamic percentage is the proportion of the operating time of the semiconductor device in the dynamic power consumption state to the total operating time of the semiconductor device in the power-on state.

[0046] In a specific embodiment of the present invention, the method for setting several operating scenarios for semiconductor devices is as follows: Three types of operating scenarios are set, including: light scenario, medium scenario, and heavy scenario; in the light scenario, the operating frequencies of the semiconductor devices include 0, 50MHz, and 100MHz, the ambient temperature data includes -40℃, 0℃, 25℃, 40℃, and 60℃, the power supply voltage of the semiconductor devices VDD=1.0V, and the dynamic ratio is 0.3; in the medium scenario, the operating frequencies of the semiconductor devices include 200MHz, 50MHz, and 100MHz. For GHz, 500GHz, 800GHz, and 1GHz, ambient temperature data includes 25℃, 45℃, 65℃, and 85℃, and the power supply voltage VDD of semiconductor devices includes 1.0V and 1.1V, with a dynamic ratio of 0.5. In heavy-duty scenarios, the operating frequencies of semiconductor devices include 1.5GHz, 2GHz, 2.5GHz, and 3GHz, ambient temperature data includes 60℃, 80℃, 100℃, and 125℃, and the power supply voltage VDD of semiconductor devices is 1.2V, with a dynamic ratio of 0.8.

[0047] Then, the value range corresponding to the design parameters of the semiconductor device is set, and several design parameters of the semiconductor device are determined by uniformly sampling within the value range corresponding to the design parameters; the semiconductor device under any working scenario and any design parameters is simulated, and the corresponding static and dynamic simulation data are collected respectively; the output data is obtained through scripts and data interfaces of different platforms; the static power consumption is collected through HSPICE DC simulation (DC mode); the dynamic power consumption is collected through HSPICE transient simulation (TRAN mode), and the simulation duration is set to 10 signal cycles to ensure data stability.

[0048] In one specific embodiment of the present invention, the threshold voltage of the semiconductor device is used as the design parameter of the semiconductor device, and the value range of the threshold voltage is set to be 0.3V-0.6V, and the sampling step size within the value range is set to 0.05V. In other embodiments, it can be adjusted according to the actual situation. The embodiments of the present invention do not impose specific limitations.

[0049] Finally, the operating data of the semiconductor device under various working scenarios is obtained through the simulation platform. The operating data includes junction temperature data, current data and ambient temperature data of the semiconductor device, and the collected operating data is preprocessed.

[0050] Specifically, for the aforementioned simulation platform, in this embodiment of the invention, simulation data analysis is performed by combining ANSYS Icepak simulation software, HSPICE simulation software, and Synopsys Prime Power simulation software.

[0051] It should be noted that ANSYS Icepak simulation software is an existing electronic thermal analysis software, HSPICE simulation software is a circuit simulation software that supports DC analysis, transient analysis and frequency domain analysis, and Synopsys Prime Power simulation software is a power analysis software for integrated circuit design that supports static power consumption and dynamic power consumption. Since the simulation process of ANSYS Icepak simulation software, HSPICE simulation software and Synopsys Prime Power simulation software for components is an existing method, this embodiment of the invention will not elaborate on it.

[0052] As an optional embodiment, the preprocessing of the collected working data includes the following specific methods: filtering the current data using a Gaussian filtering algorithm, and then performing Z-Score normalization on the filtered current data.

[0053] Thus, the above methods have yielded several working scenarios and the working data of semiconductor devices in each scenario.

[0054] Step S002: Based on the difference between the junction temperature data of the semiconductor device and the ambient temperature data under any working scenario, and combined with the dynamic proportion under the working scenario, the temperature sensitivity coefficient of the semiconductor device under the working scenario is obtained.

[0055] It's important to note that traditional power consumption simulation methods only analyze simulation data and iterate parameters for a single mode. The simulation results are specific to a particular scenario and deviate significantly from actual power consumption during use. This leads to discrepancies between optimized parameters and real-world applications. Furthermore, semiconductor devices typically alternate between active and idle states. The dynamic power consumption of semiconductor devices varies under different operating conditions, resulting in different actual energy consumption and heat accumulation levels. Moreover, during operation, heat accumulation causes changes in ambient temperature, altering the leakage current carrier mobility characteristics and thus impacting the relationship between design parameters and power consumption.

[0056] As a preferred embodiment, the specific method for obtaining the temperature sensitivity coefficient is as follows:

[0057] First, any working scenario is designated as the target working scenario. The dynamic proportion of the target working scenario is obtained. Then, the dynamic temperature difference coefficient of the semiconductor device under the target working scenario is calculated by combining the difference between the junction temperature data of the semiconductor device and the ambient temperature data under the target working scenario.

[0058] It's important to note that in actual semiconductor device operation, dynamic power consumption generates heat, leading to a rise in junction temperature. This temperature increase alters the leakage current and carrier mobility characteristics of the semiconductor device, resulting in deviations in static power consumption. The junction temperature data reflects the actual operating temperature inside the semiconductor device during operation and is typically higher than the ambient temperature. A greater difference between the junction temperature and the ambient temperature indicates a larger amount of heat generated during operation. The more heat accumulates, the greater the temperature change, leading to greater changes in the semiconductor's static characteristics and ultimately, a larger change in static power consumption. Therefore, the dynamic power consumption temperature difference is calculated first. Considering the varying proportions of dynamic and static power consumption in different scenarios, a correction factor is needed to calculate the dynamic temperature difference coefficient, taking into account the dynamic proportion in different scenarios.

[0059] As an optional embodiment, the specific method for obtaining the dynamic temperature difference coefficient is as follows: the difference between the junction temperature data of the semiconductor device and the ambient temperature under the target working scenario is recorded as the temperature difference factor of the semiconductor device. The temperature difference factor is corrected by the dynamic proportion under the target working scenario to obtain the dynamic temperature difference coefficient of the semiconductor device under the target working scenario. The dynamic proportion and the temperature difference factor are both positively correlated with the dynamic temperature difference coefficient.

[0060] As an optional embodiment, the specific calculation method for the dynamic temperature difference coefficient can be:

[0061]

[0062] in, This represents the dynamic temperature coefficient of a semiconductor device under the target operating environment; This indicates the junction temperature data of a semiconductor device. This represents the ambient temperature data under the target working environment; This indicates the dynamic proportion of the target work scenario.

[0063] The dynamic energy consumption temperature difference mentioned above represents the temperature change of the semiconductor device under this parameter combination. Therefore, the static energy consumption change of the semiconductor device under this temperature change can be determined by this temperature change.

[0064] Then, a reference working scenario is selected from all working scenarios based on the dynamic temperature difference coefficient of the semiconductor device in the target working scenario and the ambient temperature coefficient in the target working scenario.

[0065] As an optional embodiment, the specific method for obtaining the reference working scenario is as follows: combining the ambient temperature data under the target working scenario and the dynamic temperature difference coefficient of the semiconductor device under the target working scenario, the comprehensive temperature of the semiconductor device under the target working scenario is obtained; the ambient temperature data corresponding to the smallest absolute value of the difference between the ambient temperature data corresponding to all working scenarios and the comprehensive temperature is obtained, and recorded as the reference ambient temperature; the working scenario to which the reference ambient temperature belongs is obtained, and the working scenario is consistent with all other scenario parameters in the target working scenario except for the ambient temperature data, and the working scenario is recorded as the reference working scenario.

[0066] As an optional embodiment, the specific method for calculating the overall temperature of the semiconductor device under the target operating scenario can be: ,in This indicates the overall temperature of the semiconductor device under the target operating environment; This represents the ambient temperature data under the target working environment; It represents the dynamic temperature difference coefficient of a semiconductor device under the target operating environment.

[0067] Since the temperature of semiconductor devices varies under different operating scenarios and design parameters, their static power consumption also varies. This reflects the degree of temperature change caused by power consumption changes in different operating scenarios and design parameters. In order to quantify this degree of temperature change more effectively, this embodiment of the invention utilizes the difference in static power consumption of semiconductor devices under different ambient temperature conditions and combines it with the dynamic temperature difference coefficient to quantify the temperature sensitivity coefficient of semiconductor devices.

[0068] Finally, the static power consumption data of the semiconductor device under the target operating scenario is obtained, and the static power consumption data of the semiconductor device under the reference operating scenario is recorded as the reference static power consumption data of the semiconductor device. Based on the difference between the static power consumption data of the semiconductor device under the target operating scenario and the reference static power consumption data, and combined with the dynamic temperature difference coefficient of the semiconductor device under the target operating scenario, the temperature sensitivity coefficient of the semiconductor device under the target operating scenario is calculated.

[0069] As an optional embodiment, the specific method for calculating the temperature sensitivity coefficient can be:

[0070]

[0071] in, This indicates the temperature sensitivity coefficient of a semiconductor device under the target operating environment; Represents reference static power consumption data for semiconductor devices; This represents the static power consumption data of the semiconductor under the target operating scenario; This represents the dynamic temperature coefficient of a semiconductor device under the target operating environment; This represents the preset hyperparameters.

[0072] It should be noted that hyperparameters are set in the embodiments of the present invention. The value can be 0.1 to avoid a denominator of 0. In other embodiments, adjustments can be made based on actual conditions; this embodiment of the invention does not impose specific limitations. The temperature sensitivity coefficient describes the sensitivity of a semiconductor device's static power consumption to temperature under a target operating scenario. The larger the temperature sensitivity coefficient, the greater the change in the semiconductor device's static power consumption when the ambient temperature changes under the target operating scenario. In light usage scenarios, ambient temperature is the primary factor, heat generation is low, and temperature has a weak impact on static power consumption, resulting in a smaller temperature sensitivity coefficient. In heavy usage scenarios, heat generation is significant, and temperature has a strong impact, resulting in a larger temperature sensitivity coefficient. Therefore, different temperature sensitivity thresholds are set for different scenarios. For sensitivity exceeding the threshold, subsequent static power consumption correction is performed.

[0073] Thus, the temperature sensitivity coefficient of semiconductor devices under different operating scenarios has been obtained through the above method.

[0074] Step S003: Adjust the static power consumption data of the semiconductor device under different design parameters using the temperature sensitivity coefficient, and perform weighted processing on the adjustment results of the static power consumption data change and the dynamic power consumption data change using the dynamic ratio in the working scenario to obtain the weighted static power consumption change and weighted dynamic power consumption change of the semiconductor device under each design parameter in the working scenario.

[0075] It's important to note that traditional methods for optimizing design parameters through simulation data analysis typically consider only individual modes. Optimizing design parameters under both static and dynamic power consumption modes can lead to conflicting results. That is, optimizing parameters under different power consumption modes will inevitably result in different optimization directions. Therefore, it's necessary to consider the specific application scenario of the semiconductor device and the dominance of different power consumption modes within that scenario when optimizing design parameters.

[0076] As a preferred embodiment, the specific method for obtaining the weighted static power consumption change and the weighted dynamic power consumption change is as follows:

[0077] First, data interpolation is performed on the static power consumption data and dynamic power consumption data corresponding to all design parameters of the semiconductor device under the target operating environment to obtain the second design-static power consumption sequence and the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment.

[0078] As an optional embodiment, the specific method for obtaining the second design-static power consumption sequence and the second design-dynamic power consumption sequence is as follows: obtain a sequence of static power consumption data corresponding to all design parameters of the semiconductor device under the target operating environment, denoted as the first design-static power consumption sequence of the semiconductor device under the target operating environment; perform interpolation processing on the first design-static power consumption sequence using the least squares method to obtain the second design-static power consumption sequence of the semiconductor device under the target operating environment; replace the static power consumption data in the second design-static power consumption sequence with dynamic power consumption data to obtain the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment.

[0079] It should be noted that increasing the threshold voltage of a semiconductor device reduces static leakage current and optimizes static power consumption; however, it also increases the switching delay of the semiconductor device, increasing dynamic power consumption. Therefore, further analysis of the static and dynamic power consumption of the semiconductor device is needed to determine the power consumption variation under different design parameters. Furthermore, since the design parameters of semiconductor devices are typically discrete data, to enable a more detailed analysis of the static and dynamic power consumption, this embodiment of the invention interpolates the static and dynamic power consumption data under different design parameters during the semiconductor operation in the target working environment. This increases the amount of static and dynamic power consumption data, facilitating subsequent power consumption sequence analysis and improving the effectiveness of optimizing the semiconductor device's design parameters during power consumption simulation. Specifically, when multiple design parameters exist during the semiconductor operation in other embodiments, the steps of obtaining the static and dynamic power consumption sequences are repeated to obtain the static and dynamic power consumption sequences under each design parameter.

[0080] Then, based on the differences in static power consumption data and dynamic power consumption data between any design parameter and adjacent design parameters in the second design-static power consumption sequence and the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment, the static power consumption change and dynamic power consumption change of the corresponding design parameters of the semiconductor device under the target operating environment are determined.

[0081] As an optional embodiment, the specific method for obtaining the static power consumption change and the dynamic power consumption change is as follows: As a semiconductor device, in the second design-static power consumption sequence, The static power consumption variation of each design parameter, where This indicates the semiconductor device in the second design-static power consumption sequence. Static power consumption data for each design parameter. This indicates the semiconductor device in the second design-static power consumption sequence. The static power consumption data of each design parameter; the second design-static power consumption sequence in the method for obtaining the static power consumption change is replaced with the second design-dynamic power consumption sequence, thereby obtaining the semiconductor device in the second design-dynamic power consumption sequence. The dynamic power consumption change of each design parameter.

[0082] Finally, the static power consumption data of the semiconductor device under arbitrary design parameters is adjusted and corrected by using the difference between junction temperature data and ambient temperature data and the temperature sensitivity coefficient, so as to obtain the corrected static power consumption change. The dynamic proportion under the target working environment is used as the weight to weight the dynamic power consumption change and the corrected static power consumption change, respectively, so as to obtain the weighted static power consumption change and weighted dynamic power consumption change of the semiconductor device under the target working scenario and with the target design parameters as the design parameters.

[0083] As an optional embodiment, the specific method for obtaining the corrected static power consumption change is as follows: Design parameters of arbitrary size are denoted as target design parameters. The difference between the junction temperature data of the semiconductor device and the ambient temperature data corresponding to the target operating environment is obtained when the design parameters are the target design parameters under the target operating environment. This difference is denoted as the temperature variable factor of the semiconductor device under the target operating environment and with the target design parameters as the design parameters. A correction factor is obtained by combining the temperature variable factor and the temperature sensitivity coefficient. The correction factor is used to correct the static power consumption change of the semiconductor device under the target design parameters and with the target design parameters as the design parameters, resulting in the corrected static power consumption change of the semiconductor device under the target operating environment and with the target design parameters as the design parameters. The correction factor is negatively correlated with the corrected static power consumption change, and the static power consumption change is positively correlated with the corrected static power consumption change.

[0084] Furthermore, regarding the difference between the junction temperature data of the semiconductor device and the ambient temperature data corresponding to the target operating environment, in a specific embodiment of the present invention, the difference is the difference between the junction temperature data and the ambient temperature data.

[0085] As an optional embodiment, the specific calculation method for the corrected static power consumption change can be:

[0086]

[0087] In the formula, This represents the corrected static power consumption change of a semiconductor device under the target operating environment, when the target design parameters are used as the design parameters. This indicates the temperature sensitivity coefficient of a semiconductor device under the target operating environment; This represents the temperature variable factor of a semiconductor device under the target operating environment and with the target design parameters as the design parameters. It represents the change in static power consumption of a semiconductor device under the target operating environment when the target design parameters are used as design parameters.

[0088] In addition, a preset temperature sensitivity threshold is set. When the temperature sensitivity coefficient of the semiconductor device in the target working scenario is less than the temperature sensitivity threshold, the correction of static power consumption change is skipped. Instead, the dynamic power consumption change and static power consumption change of the semiconductor device in the target working scenario and with the target design parameters as the design parameters are weighted according to the dynamic proportion of the target working scenario, so as to obtain the corresponding weighted static power consumption change and weighted dynamic power consumption change.

[0089] It should be noted that, in the actual operation of semiconductor devices, the two power consumption modes may influence each other. Therefore, this embodiment of the invention combines the temperature sensitivity coefficient and the temperature variable factor to obtain a correction factor. This corrects for changes in static power consumption, making it more closely reflect real-world conditions. Furthermore, regarding the temperature sensitivity coefficient of semiconductor devices in the target operating scenario, once the temperature sensitivity coefficient exceeds the temperature sensitivity threshold, a larger value indicates a greater change in power consumption in response to temperature variations, thus requiring a higher degree of adjustment to the static power consumption. In this embodiment of the invention, the temperature sensitivity threshold is set according to the operating scenario, specifically as follows: when the target operating scenario is a mild scenario, the temperature sensitive area is... When the target working environment is a moderate environment, the temperature-sensitive area is: When the target working environment is a heavy-duty environment, the temperature-sensitive area is: The embodiments of the present invention are not specifically limited and can be adjusted according to the actual situation.

[0090] It's important to further clarify that when optimizing the design parameters of semiconductor devices, the dominance of static and dynamic power consumption varies depending on the operating scenario. For example, in light-duty scenarios, static power consumption dominates, while in heavy-duty scenarios, dynamic power consumption dominates. Therefore, the need and state for balancing these two power consumption modes differ under different operating scenarios. Consequently, it's necessary to weight the power consumption variations under both dynamic and static power consumption modes according to the specific operating scenario, thus adapting the device to the appropriate operating condition.

[0091] As an optional embodiment, the specific calculation method for the weighted static power consumption change and the weighted dynamic power consumption change can be as follows:

[0092]

[0093] in, It represents the weighted static power consumption change of a semiconductor device under the target operating scenario and with the target design parameters as the design parameters; It represents the weighted dynamic power consumption change of a semiconductor device under the target operating scenario and with the target design parameters as the design parameters; Indicates the dynamic proportion within the target work scenario; This represents the corrected static power consumption change of a semiconductor device under the target operating environment, when the target design parameters are used as the design parameters. It represents the dynamic power consumption change of a semiconductor device under the target operating environment when the target design parameters are used as design parameters.

[0094] Thus, the weighted static power consumption change and the weighted dynamic power consumption change are obtained through the above methods.

[0095] Step S004: Based on the difference between the weighted static power consumption change and the weighted dynamic power consumption change, the design parameters of the semiconductor device are iteratively optimized to obtain the optimal design parameters of the semiconductor device.

[0096] As a preferred embodiment, the specific method for obtaining the optimal design parameters of the semiconductor device is as follows:

[0097] First, based on the difference between the weighted static power consumption change and the weighted dynamic power consumption change of the semiconductor device under the target operating scenario and with the target design parameters as the design parameters, the power consumption balance coefficient of the semiconductor device under the target operating scenario and with the target design parameters as the design parameters is calculated.

[0098] As an optional embodiment, the specific calculation method for the power consumption balance coefficient can be:

[0099]

[0100] in, It represents the power consumption balance coefficient of a semiconductor device under the target operating scenario and with the target design parameters as the design parameters; It represents the weighted static power consumption change of a semiconductor device under the target operating scenario and with the target design parameters as the design parameters; It represents the weighted dynamic power consumption change of a semiconductor device under the target operating scenario and with the target design parameters as the design parameters; This represents the absolute value function.

[0101] It should be noted that when the design parameters of a semiconductor device change, the static power consumption and dynamic power consumption change in opposite directions. Therefore, when the design parameters change, the closer the absolute values ​​of the changes in static power consumption and dynamic power consumption are, the closer the changes in the two power consumptions are. At this time, the design parameters are more likely to reach a balance between the two states. That is, when the power consumption balance coefficient K is closer to 1, it means that the current design parameters are closer to the balance between the two power consumptions in this working scenario. Then, the design parameters are the optimal design parameters for the semiconductor device in this working scenario.

[0102] Then, the design parameters of the semiconductor device under the target operating scenario are iteratively updated to obtain the power consumption balance coefficient of each design parameter of the semiconductor device under the target operating scenario; a qualified range is preset, and when the value of the power consumption balance coefficient is within the qualified range, the design parameter corresponding to the power consumption balance coefficient is taken as the optimal design parameter of the semiconductor device under the target operating scenario.

[0103] It should be noted that when iteratively updating the design parameters of semiconductor devices, once the power balance coefficient corresponding to the design parameters begins to fall within the acceptable range, the subsequent update iteration count is set to 5 times to avoid over-optimization leading to a surge in design complexity. The preset update iteration count can be adjusted according to actual conditions, and this embodiment of the invention does not impose a specific limitation. Furthermore, in this embodiment of the invention, the acceptable range can be preset to... The embodiments of the present invention are not specifically limited and can be adjusted according to the actual situation.

[0104] Through the above steps, the power consumption simulation analysis and design parameter optimization of semiconductor devices are completed.

[0105] Please see Figure 2 The present invention also provides a semiconductor device power consumption simulation data analysis system, including a processor 201, a memory 202, and a program 2021 stored in the memory 202 and executable on the processor 201.

[0106] When program 2021 is executed by processor 201, it can achieve the following: Figure 1 Any steps in the corresponding method embodiments and the achievement of the same beneficial effects will not be repeated here.

[0107] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by hardware related to program instructions, and the program can be stored in a readable medium.

[0108] This invention also provides a readable storage medium storing a computer program, which, when executed by a processor, can perform the above-described functions. Figure 1 Any step in the corresponding method embodiment can achieve the same technical effect, and will not be repeated here to avoid repetition.

[0109] The computer-readable storage medium of this invention can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. For example, a computer-readable storage medium can be an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0110] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0111] The program code contained on the storage medium can be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0112] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or terminal. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0113] This invention also provides a computer program product that, when run on a computer, causes the computer to perform the aforementioned steps to implement the semiconductor device power consumption simulation data analysis method provided in the above embodiments.

[0114] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for analyzing power consumption simulation data of semiconductor devices, characterized in that, The method includes the following steps: Several operating scenarios and design parameters of semiconductor devices are set, and the semiconductor devices under each operating scenario and design parameter are simulated to obtain the dynamic proportion of the operating scenario, junction temperature data, ambient temperature data, static power consumption data, and dynamic power consumption data of the semiconductor devices during the simulation process. Based on the difference between junction temperature data and ambient temperature data of semiconductor devices under any working scenario, and combined with the dynamic proportion under the working scenario, the temperature sensitivity coefficient of semiconductor devices under the working scenario is obtained. The static power consumption data of semiconductor devices under different design parameters is adjusted by using a temperature sensitivity coefficient. The adjustment results of the static power consumption data and the dynamic power consumption data under the working scenario are weighted and processed to obtain the weighted static power consumption change and weighted dynamic power consumption change of semiconductor devices under each design parameter in the working scenario. The design parameters of the semiconductor device are iteratively optimized based on the difference between the weighted static power consumption change and the weighted dynamic power consumption change to obtain the optimal design parameters of the semiconductor device under the stated working scenario. The specific method for obtaining the temperature sensitivity coefficient is as follows: Any working scenario is recorded as the target working scenario. The dynamic proportion of the target working scenario is obtained. The dynamic temperature difference coefficient of the semiconductor device in the target working scenario is calculated by combining the difference between the junction temperature data of the semiconductor device and the ambient temperature data in the target working scenario. Based on the dynamic temperature difference coefficient of the semiconductor device in the target working scenario and the ambient temperature coefficient in the target working scenario, a reference working scenario is selected from all working scenarios. Acquire the static power consumption data of the semiconductor device in the target working scenario, record the static power consumption data of the semiconductor device in the reference working scenario as the reference static power consumption data of the semiconductor device, and calculate the temperature sensitivity coefficient of the semiconductor device in the target working scenario based on the difference between the static power consumption data of the semiconductor device in the target working scenario and the reference static power consumption data, and in combination with the dynamic temperature difference coefficient of the semiconductor device in the target working scenario. The specific method for obtaining the optimal design parameters is as follows: Design parameters of any size are denoted as target design parameters. Based on the difference between the weighted static power consumption change and the weighted dynamic power consumption change of the semiconductor device under the target operating scenario and with the target design parameters as the design parameters, the power consumption balance coefficient of the semiconductor device under the target operating scenario and with the target design parameters as the design parameters is calculated. The design parameters of the semiconductor device under the target operating scenario are iteratively updated to obtain the power consumption balance coefficient of each design parameter of the semiconductor device under the target operating scenario; a qualified range is preset, and when the value of the power consumption balance coefficient is within the qualified range, the design parameter corresponding to the power consumption balance coefficient is taken as the optimal design parameter of the semiconductor device under the target operating scenario.

2. The semiconductor device power consumption simulation data analysis method according to claim 1, characterized in that, The specific method for obtaining the dynamic temperature difference coefficient is as follows: The difference between the junction temperature data of the semiconductor device and the ambient temperature under the target working scenario is denoted as the temperature difference factor of the semiconductor device. The temperature difference factor is corrected by the dynamic proportion under the target working scenario to obtain the dynamic temperature difference coefficient of the semiconductor device under the target working scenario. The dynamic proportion and the temperature difference factor are both positively correlated with the dynamic temperature difference coefficient.

3. The semiconductor device power consumption simulation data analysis method according to claim 1, characterized in that, The specific method for obtaining the reference work scenario is as follows: Any work scenario includes several scenario parameters; By combining the ambient temperature data under the target working scenario and the dynamic temperature difference coefficient of the semiconductor device under the target working scenario, the comprehensive temperature of the semiconductor device under the target working scenario is obtained. The ambient temperature data with the smallest absolute value of the difference between the ambient temperature data corresponding to all working scenarios and the comprehensive temperature is obtained and recorded as the reference ambient temperature. The working scenario to which the reference ambient temperature belongs is obtained, and the working scenario is consistent with all other scenario parameters in the target working scenario except for the ambient temperature data. The working scenario is recorded as the reference working scenario.

4. The semiconductor device power consumption simulation data analysis method according to claim 1, characterized in that, The specific methods for obtaining the weighted static power consumption change and the weighted dynamic power consumption change are as follows: Data interpolation is performed on the static power consumption data and dynamic power consumption data corresponding to all design parameters of the semiconductor device under the target operating environment to obtain the second design-static power consumption sequence and the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment. Based on the differences in static power consumption data and dynamic power consumption data between any design parameter and adjacent design parameters in the second design-static power consumption sequence and the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment, determine the static power consumption change and dynamic power consumption change of the corresponding design parameter of the semiconductor device under the target operating environment. By utilizing the difference between junction temperature data and ambient temperature data, as well as the temperature sensitivity coefficient, the static power consumption data of semiconductor devices under arbitrary design parameters is adjusted and corrected to obtain the corrected static power consumption change. Using the dynamic proportion under the target operating environment as the weight, the dynamic power consumption change and the corrected static power consumption change are weighted separately to obtain the weighted static power consumption change and the weighted dynamic power consumption change of the semiconductor device under the target operating scenario and with the target design parameters as the design parameters.

5. The semiconductor device power consumption simulation data analysis method according to claim 4, characterized in that, The specific methods for obtaining the second design-static power consumption sequence and the second design-dynamic power consumption sequence are as follows: A sequence of static power consumption data corresponding to all design parameters of a semiconductor device under a target operating environment is obtained and denoted as the first design-static power consumption sequence of the semiconductor device under the target operating environment. The first design-static power consumption sequence is interpolated using the least squares method to obtain the second design-static power consumption sequence of the semiconductor device under the target operating environment. The static power consumption data in the second design-static power consumption sequence is replaced with dynamic power consumption data to obtain the second design-dynamic power consumption sequence of the semiconductor device under the target operating environment.

6. The semiconductor device power consumption simulation data analysis method according to claim 4, characterized in that, The specific methods for obtaining the static power consumption change and the dynamic power consumption change are as follows: Will As a semiconductor device, in the second design-static power consumption sequence, The static power consumption variation of each design parameter, where This indicates the semiconductor device in the second design-static power consumption sequence. Static power consumption data for each design parameter. This indicates the semiconductor device in the second design-static power consumption sequence. The static power consumption data of each design parameter; the second design-static power consumption sequence in the method for obtaining the static power consumption change is replaced with the second design-dynamic power consumption sequence, thereby obtaining the semiconductor device in the second design-dynamic power consumption sequence. The dynamic power consumption change of each design parameter.

7. The semiconductor device power consumption simulation data analysis method according to claim 4, characterized in that, The specific method for obtaining the corrected static power consumption change is as follows: The difference between the junction temperature data of the semiconductor device and the ambient temperature data corresponding to the target operating environment is obtained when the design parameters are the target operating environment and the design parameters are the target design parameters. This difference is denoted as the temperature variable factor of the semiconductor device when the design parameters are the target operating environment and the design parameters are the target design parameters. A correction factor is obtained by combining the temperature variable factor and the temperature sensitivity coefficient. The correction factor is used to correct the static power consumption change of the semiconductor device under the target design parameters and with the target design parameters as the design parameters. The corrected static power consumption change of the semiconductor device under the target operating environment with the target design parameters as the design parameters is obtained. The correction factor is negatively correlated with the corrected static power consumption change, and the static power consumption change is positively correlated with the corrected static power consumption change.

8. A semiconductor device power consumption simulation data analysis system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the semiconductor device power consumption simulation data analysis method as described in any one of claims 1 to 7.

Images