A method and device for serdes chip performance control parameter optimization

Abstract

The application provides a serdes chip performance control parameter optimization method and device, comprising: setting a value range of all configuration parameters of a preset configuration type in a serdes chip, generating an initial particle swarm according to each configuration parameter and a corresponding value range; inputting the initial particle swarm as an input particle swarm into a first iteration; each iteration comprises: inputting each particle in the input particle swarm into a corresponding serdes chip, obtaining an eye diagram corresponding to each particle, and obtaining a target value corresponding to each particle according to the eye diagram corresponding to each particle; obtaining a maximum target value in all particles in the input particle swarm as a local optimal value, and obtaining a local optimal value with the maximum target value in all iteration rounds as a global optimal value; after a termination condition is met, an optimal configuration parameter of the serdes chip is obtained, and the optimization efficiency and the channel compensation performance are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of signal transmission, in particular to a serdes chip performance control parameter optimization method and device. BACKGROUND

[0002] With the acceleration of data center and artificial intelligence center construction and the increasingly widespread demand for high-performance computing, the communication bandwidth requirements between system chips and cabinets are growing exponentially. Compared with parallel communication, serial communication (serdes) has significant advantages in link cost, data rate and anti-interference ability, and is widely used in communication systems. Since 2015, the speed of high-speed serial communication has increased rapidly, almost doubling every 3 to 5 years. The Peripheral Component Interconnect Express (PICE) has developed from the first generation supporting a transmission rate of 2.5 GB / s to the sixth generation supporting a transmission rate of 112 GB / s. However, the serdes system is a complex digital-analog signal mixed system, with higher transmission rates and more serious signal attenuation on the channel, which requires compensation for signal attenuation on the channel. It can be said that the use of equalization technology for channel compensation of the serdes serial communication system is the core of the serdes serial communication system.

[0003] The existing serdes transmission system such as Figure 1As shown, parallel data is sent to the multiplexer (MUX), and the parallel data is first converted into serial data by the multiplexer (MUX). Since the signal will have different attenuation and performance loss after passing through the channel, pre-emphasis is performed by the feed forward equalizer (FFE) at the transmitting end to balance the high and low frequency components of the signal and compensate for the loss of part of the channel. During the signal link transmission process, high frequency loss is more serious. At the receiving end, the continuous time linearlity equalizer (CTLE) is usually used to compensate for the high frequency attenuation of the signal from the frequency domain, thereby eliminating part of the pre- and post- tags in the signal, and then the variable gain amplifier (VGA) is used to further amplify the signal amplitude. This process also amplifies crosstalk and noise, so the feedback decision equalizer (DFE) is usually used at the receiving end to eliminate most of the post-tags. Finally, the high-speed serial data is converted into low-speed parallel data by the demultiplexer (DeMUX). The sampler (Sampler) is used to convert the continuous analog signal into a digital signal, and the clock data recovery circuit (CDR) and the phase interpolator (PI) are used to continuously adjust the optimal sampling point of the clock lock signal. The phase locked loop (PLL) is used to achieve locking with the external reference clock.

[0004] At present, the channel compensation mode of the serdes system is mostly realized by adjusting one or more of the parameters of the continuous-time linear equalizer, the parameters of the amplifier, the tap coefficients of the feedback decision equalizer and the tap coefficients of the discrete-time linear equalizer. In the process of backsheet testing or application, a set of parameters of the continuous-time linear equalizer, the amplifier, the discrete-time linear equalizer and the feedback decision equalizer need to be found to obtain good link transmission performance. However, if the actual situation of the channel (such as the attenuation size and whether the channel design meets the standard requirements) cannot be known or the engineer is not familiar with the serdes link architecture, the 10-level continuous-time linear equalizer, the 10-level amplifier, the discrete-time linear equalizer with 6 taps (6 taps represent that the discrete-time linear equalizer can solve the tailing influence of the first 6 symbols on the current symbol) are used for optimization, each tap is 5 bits, the feedback decision equalizer is configured with 3 taps, each tap has a bit width of 5 bits, and there are about 1.85 million combinations. Under the existing manual debugging optimization method, it is extremely difficult to find a suitable set of parameters, the efficiency is low, and it cannot better adapt to the current testing requirements.

[0005] Therefore, overcoming the defects of the prior art is a problem to be solved in the technical field. SUMMARY

[0006] The technical problem to be solved by the present application is how to improve the debugging efficiency in the process of debugging the serdes system and compensating the serial communication channel.

[0007] The present application adopts the following technical solutions:

[0008] In a first aspect, a method for optimizing performance control parameters of a serdes chip is provided, comprising:

[0009] Setting the value range of all configuration parameters of a preset configuration type in the serdes chip, and generating an initial particle swarm according to each configuration parameter and the corresponding value range;

[0010] Inputting the initial particle swarm as an input particle swarm into the first iteration;

[0011] Each iteration includes: inputting each particle in the input particle swarm into the corresponding serdes chip, obtaining the eye diagram corresponding to each particle, and obtaining the target value corresponding to each particle according to the eye diagram corresponding to each particle; obtaining the maximum target value of each particle in the historical iteration round as the local optimal value of the particle, and obtaining the maximum target value of all particles in the current iteration round as the global optimal value; and determining whether the termination condition is met;

[0012] When the termination condition is met, the configuration parameters in the particle corresponding to the current global optimal value are used as the optimal configuration.

[0013] When the termination condition is not satisfied, each particle in the current input particle swarm is iteratively updated according to the local optimal value and the global optimal value of the particle, to obtain an input particle swarm for the next iteration and input into the next iteration.

[0014] Preferably, the target value corresponding to each particle is obtained according to the eye diagram corresponding to the particle, and specifically includes:

[0015] ;

[0016] wherein, the target value, an eye height in the eye diagram, a weight of the eye height, an eye width in the eye diagram, a weight of the eye width.

[0017] Preferably, the preset configuration type includes one or more of a continuous-time linear equalizer parameter, an amplifier parameter, a tap coefficient of a feedback decision equalizer, a proportional path parameter, an integral path parameter, and a tap coefficient of a discrete-time linear equalizer.

[0018] Preferably, the value range of all configuration parameters of the preset configuration type in the serdes chip is set, and specifically includes:

[0019] The value range of the continuous-time linear equalizer parameter is: wherein, a parameter bit width of the continuous-time linear equalizer parameter;

[0020] The value range of the amplifier parameter is: wherein, a parameter bit width of the continuous-time linear equalizer parameter;

[0021] The value range of the tap coefficient of the feedback decision equalizer is: wherein, a bit width of the i-th tap coefficient of the feedback decision equalizer;

[0022] The value range of the proportional path parameter is: wherein, a parameter bit width of the proportional path parameter;

[0023] The value range of the integral path parameter is: wherein, a parameter bit width of the integral path parameter;

[0024] The range of values of each tap coefficient of the discrete-time linear equalizer is: wherein is a bit width of the i-th tap coefficient of the discrete-time linear equalizer.

[0025] Preferably, the method further comprises the following steps of:

[0026] each particle in the current iteration obtains an update difference corresponding to itself according to the local optimal value of itself and the global optimal value;

[0027] each particle in the current iteration is updated according to the update difference corresponding to itself to obtain a particle corresponding to the next iteration, and then an input particle swarm of the next iteration is obtained.

[0028] Preferably, the method further comprises the following steps of:

[0029] the formula of the update difference is:

[0030] ;

[0031] wherein, is a particle update difference in the n-th iteration, is a particle update difference in the n+1-th iteration, and is a random number with a value range of is a random number with a value range of is a particle corresponding to the local optimal value of the current iteration, is a particle corresponding to the current global optimal value, is the i-th particle in the current iteration. Preferably, the termination condition comprises the following steps of:

[0032] setting a preset iteration number;

[0033] judging whether the current iteration round is greater than or equal to the preset iteration number;

[0034] when the current iteration round is greater than or equal to the preset iteration number, the termination condition is satisfied;

[0035] when the current iteration round is less than the preset iteration number, the termination condition is not satisfied.

[0036] when the current iteration round is less than the preset iteration number, the termination condition is not satisfied.

[0037] ​Preferably, the initial update difference of the particles in the initial particle group is calculated as follows:

[0038] ;

[0039] wherein, is the initial update difference of the i-th particle in the initial particle group, is the particle with all parameters being maximum in the particle, is the particle with all parameters being minimum in the particle, is the preset iteration number.

[0040] In a second aspect, a device for serdes chip performance control parameter optimization is provided, comprising at least one processor, and a memory connected with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the processor to execute the method for serdes chip performance control parameter optimization.

[0041] In a third aspect, the present application further provides a non-volatile computer storage medium, which stores computer executable instructions, and the computer executable instructions are executed by one or more processors to complete the method in the first aspect.

[0042] In a fourth aspect, a chip is provided, comprising a processor and an interface, which is used to call and run a computer program stored in a memory to execute the method in the first aspect.

[0043] In a fifth aspect, a computer program product containing instructions is provided, which makes a computer or a processor execute the method in the first aspect when the instructions are executed on the computer or the processor.

[0044] In a sixth aspect, a system for serdes chip performance control parameter optimization is provided, which comprises the device for serdes chip performance control parameter optimization in the second aspect and uses the method for serdes chip performance control parameter optimization in the first aspect.

[0045] The application provides a serdes chip performance control parameter optimization method and device, which comprises the following steps: setting the value range of all configuration parameters of a preset configuration type in a serdes chip; generating an initial particle swarm according to each configuration parameter and the corresponding value range; inputting the initial particle swarm into the first iteration as an input particle swarm; in each iteration, the following steps are included: inputting each particle in the input particle swarm into the corresponding serdes chip, obtaining the eye diagram corresponding to each particle, and obtaining the target value corresponding to each particle according to the eye diagram; obtaining the maximum target value in all particles in the input particle swarm as a local optimal value, and obtaining the local optimal value with the maximum target value in all iteration rounds as a global optimal value; obtaining the optimal configuration parameter of the serdes chip after the termination condition is met. The eye diagram output by the link transmission is taken as the optimization target, the configuration parameter affecting the optimization target is taken as the independent variable, the optimal configuration parameter suitable for the current link is found by using the iteration mode, and the optimization efficiency and the channel compensation performance are improved. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the application, the drawings needed to be used in the embodiments of the application will be briefly introduced. Obviously, the drawings described below are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0047] Figure 1 is a schematic diagram of an existing serdes system provided by the embodiments of the application;

[0048] Figure 2 is a method flowchart of a serdes chip performance control parameter optimization method provided by the embodiments of the application;

[0049] Figure 3 is a schematic diagram of writing and reading each particle in the corresponding serdes system in the serdes chip performance control parameter optimization method provided by the embodiments of the application;

[0050] Figure 4 is a method flowchart of particle swarm updating iteration in the serdes chip performance control parameter optimization method provided by the embodiments of the application;

[0051] Figure 5 is an iteration termination flowchart of the serdes chip performance control parameter optimization method provided by the embodiments of the application;

[0052] Figure 6 is a structural schematic diagram of a serdes chip performance control parameter optimization device provided by the embodiments of the application. DETAILED DESCRIPTION

[0053] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0054] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present disclosure and simplify the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present disclosure.

[0055] In the description of the present application, the terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the embodiments of the present disclosure, unless otherwise specified, the meaning of "a plurality of" is two or more. In addition, for example, in the description, the same type of nouns will also be described as two independent individuals by adding "A", "B" at the end, in which case the features defined with "A", "B" are only used for the purpose of distinguishing the same type of individual description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated.

[0056] "About", "approximately" or "approximately" used in the present application includes the stated value and the average value within an acceptable deviation range of the specific value, wherein the acceptable deviation range is determined by considering the measurement being discussed and the error related to the measurement of the specific quantity, i.e. the limitation of the measurement system.

[0057] Unless the context otherwise requires, throughout the specification and claims, the term "comprising" is interpreted as openly inclusive, meaning "including, but not limited to." In the description of the specification, terms such as "one embodiment," "some embodiments," "exemplary embodiment," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this disclosure. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics mentioned may be included in any suitable manner in any one or more embodiments or examples; that is, although they may be incorporated into embodiments or examples using the above terms for reasons such as order and position, it does not limit them to be incorporated in combination by a single embodiment or example.

[0058] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0059] This invention provides a method for optimizing performance control parameters of a SerDes chip, such as... Figure 2 As shown, the method flow includes:

[0060] In step 101, the value range of all configuration parameters of the preset configuration type in the serdes chip is set, and the initial particle swarm is generated according to each configuration parameter and its corresponding value range.

[0061] The method provided in this embodiment is applied to channel compensation in a SERDES system. Different parameters in the SERDES system can affect the efficiency and method of channel compensation. To achieve relatively optimal channel compensation performance and improve debugging efficiency, it is necessary to find a set of optimal parameters for the SERDES system to obtain relatively optimal link transmission performance and improve channel compensation performance. Therefore, this embodiment uses an algorithm to iterate on the corresponding parameters to obtain the optimal parameter configuration to achieve the best channel compensation performance.

[0062] The preset parameter types include one or more of the following: continuous-time linear equalizer parameters, amplifier parameters, tap coefficients of feedback decision equalizers, proportional path parameters, integral path parameters, and tap coefficients of discrete-time linear equalizers.

[0063] Based on the above, the expression for the i-th particle in the particle swarm can be:

[0064] ;

[0065] in, for the i-th particle in the input particle swarm, for a continuous-time linear equalizer parameter, for an amplifier parameter, for each tap coefficient of a feedback decision equalizer, for a proportional path parameter, for an integral path parameter, for each tap coefficient of a discrete-time linear equalizer.

[0066] The discrete-time linear equalizer is used to emphasize the signal to balance the high frequency and low frequency components of the signal, to make up for the loss of part of the signal. The continuous-time linear equalizer is used to compensate for the high frequency attenuation of the signal from the frequency domain, thereby eliminating part of the pre- and post- tags in the signal. The amplifier is used to amplify the amplitude of the signal. The feedback decision equalizer is used to eliminate most of the post-tags.

[0067] Each particle in the initial particle swarm contains all the preset parameter types, and each parameter of each particle in the initial particle swarm is randomly generated within the corresponding value range.

[0068] In step 102, the initial particle swarm is input as an input particle swarm into the first iteration.

[0069] The initial particle swarm includes a plurality of particles, and each particle is iterated.

[0070] In step 103, each iteration includes: inputting each particle in the input particle swarm into a corresponding serdes chip, obtaining an eye diagram corresponding to each particle, and obtaining a target value corresponding to each particle according to the eye diagram corresponding to each particle; obtaining the maximum target value of each particle in the historical iteration round as the local optimal value of the particle, and obtaining the maximum target value of all particles in all iteration rounds as the global optimal value; and determining whether the termination condition is met.

[0071] In the iteration process, the particles need to converge in the desired direction, so as to gradually obtain the optimal solution. In this process, the corresponding objective function needs to be determined. The target value in the objective function is used to represent the channel compensation performance and efficiency corresponding to the particle. The larger the target value, the stronger the channel compensation performance and efficiency corresponding to the particle. The smaller the target value, the weaker the channel compensation performance and efficiency corresponding to the particle. The independent variable in the objective function is obtained from the parameters in the particle.

[0072] In the embodiment, each parameter in the micro-particle can be applied to the serdes chip, the eye diagram corresponding to the micro-particle is obtained in the serdes chip, the eye height and the eye width of the eye diagram are taken as the independent variables, and the independent variables are substituted into the objective function to obtain the objective value corresponding to the micro-particle, which is used to measure the channel compensation performance and efficiency corresponding to the micro-particle.

[0073] The local optimal value is the maximum objective value in the group of each micro-particle and the historical updated micro-particle of the micro-particle. In order to more clearly explain the local optimal value, an example is given: there are 100 micro-particles in the input particle group, and the 100 micro-particles are numbered from 1 to 100 in ascending order. In each update iteration, each micro-particle is updated to become a new micro-particle, for example, the micro-particle numbered 5 is updated to become a new micro-particle numbered 5. The number of micro-particles in the input particle group in each iteration is unchanged and is 100. The local optimal value of the micro-particle numbered 5 in the current iteration is the maximum objective value corresponding to all micro-particles numbered 5 in all iteration rounds, that is, the local optimal value of the micro-particle numbered 5.

[0074] The global optimal value is the maximum objective value in all micro-particles in all iteration rounds.

[0075] In the embodiment, the eye diagram refers to a graph observed on an oscilloscope when estimating and improving the performance of a transmission system by using an experimental method. The method of observing the eye diagram is as follows: a oscilloscope is connected across the output of a receiving filter, and then the horizontal scanning period of the oscilloscope is adjusted to synchronize the horizontal scanning period of the oscilloscope with the period of a received symbol. At this time, the graph displayed on the screen of the oscilloscope is the eye diagram. The eye diagram contains rich information. The influence of inter-symbol interference and noise can be observed from the eye diagram, the overall characteristics of a digital signal are embodied, and thus the advantages and disadvantages of a transmission system can be estimated. The eye width and the eye height in the eye diagram are important parameters in the eye diagram.

[0076] In step 104, when the termination condition is met, the configuration parameters in the micro-particle corresponding to the current global optimal value are taken as the optimal configuration.

[0077] In step 105, when the termination condition is not met, each micro-particle in the current input particle group is updated according to the local optimal value and the global optimal value of the micro-particle, the input particle group in the next iteration is obtained, and the input particle group in the next iteration is input into the next iteration.

[0078] When the termination condition is not met, the next iteration needs to be continued, and the input particle swarm in the current iteration round needs to be optimized to obtain offspring. The optimization method is to converge each particle in the current iteration round to the particle corresponding to the current local optimal value and the particle corresponding to the global optimal value, so that the particle swarm in the next iteration has better performance, and the optimal solution obtained after multiple iterations can meet the relatively ideal channel compensation performance.

[0079] Further, the value range of each configuration parameter in the particle needs to be set separately. When generating the initial particle swarm, each configuration parameter in each particle in the initial particle swarm needs to be generated within the corresponding value range. When generating the offspring in each iteration round, the particles in the offspring also need to be constrained by the corresponding value range. The value range of all configuration parameters of the preset configuration type in the serdes chip is set, specifically including:

[0080] The value range of the continuous-time linear equalizer parameter is: , wherein is the parameter bit width of the continuous-time linear equalizer parameter.

[0081] The value range of the amplifier parameter is: , wherein is the parameter bit width of the continuous-time linear equalizer parameter.

[0082] The value range of each tap coefficient of the feedback decision equalizer is: , wherein is the bit width of the i-th tap coefficient of the feedback decision equalizer.

[0083] The value range of the proportional path parameter is: , wherein is the parameter bit width of the proportional path parameter.

[0084] The value range of the integral path parameter is: , wherein is the parameter bit width of the integral path parameter.

[0085] The value range of each tap coefficient of the discrete-time linear equalizer is: , wherein is the bit width of the i-th tap coefficient of the discrete-time linear equalizer.

[0086] Further, in the embodiment, the target function involves the following design:

[0087] The target value corresponding to each micro-particle is obtained according to the eye diagram corresponding to each micro-particle, and specifically includes: the formula corresponding to the target value of the eye height and the eye width is:

[0088]

[0089] wherein, the target value is, the eye height in the eye diagram is, the weight of the eye height is, the eye width in the eye diagram is, the weight of the eye width is.

[0090] Further, in the embodiment, the eye height and the eye width corresponding to each micro-particle are obtained by inputting the micro-particle into the serdes system, but considering that the number of micro-particles in the input particle group is usually large, if each micro-particle is sequentially input into a single serdes system, the optimization process will be too slow, but it is also impossible to provide a serdes system for each micro-particle, and the cost is too high to be realized. Therefore, as shown in Figure 3 , a preset number of serdes systems (L serdes systems are provided in Figure 3 ) can be provided, the preset number is less than the number of micro-particles in the input particle group and greater than 1. In actual optimization, a group of micro-particles of the preset number are input each time, the configuration parameters corresponding to the micro-particles in the group are applied to the preset number of serdes systems, respectively, the eye height and the eye width returned by each serdes system are read by the host computer, so as to obtain the eye height and the eye width corresponding to the corresponding micro-particle, and then obtain the target value corresponding to the micro-particle. The configuration parameters corresponding to the next group of micro-particles of the preset number are applied to the preset number of serdes systems, until all the micro-particles in the input particle group are applied. In the actual optimization process, the maximum time length of the eye height and the eye width returned by the serdes system is set to T milliseconds, so that the eye height and the eye width returned by each serdes system are read after T milliseconds, and the next group of eye height and eye width is input. According to the above process, until all the micro-particles in the input particle group are applied, the cost is guaranteed within the affordable range, and the optimization efficiency is improved.

[0091] Further, in the embodiment, the iteration and update of the input particle group need to ensure that the offspring generated in each iteration converges to the direction of better channel compensation efficiency, so the iteration and update need to be performed according to the local optimal value and the global optimal value. Therefore, the embodiment also relates to the following design:

[0092] Each micro-particle in the current input particle group is iteratively updated according to the local optimal value and the global optimal value of the micro-particle, to obtain an input particle group for the next iteration and input into the next iteration, as shown in​Figure 4 As shown, the method flow includes:

[0093] In step 201, each particle in this iteration obtains its corresponding update difference based on its local optimum and global optimum.

[0094] The formula for updating the difference is:

[0095] ;

[0096] in, For the i-th particle, the particle update difference in the nth iteration. This represents the particle update difference for the i-th particle in the (n+1)-th iteration. and The range of values ​​is random numbers, The range of values ​​is random numbers, Let i be the particle corresponding to the local optimum value of the i-th particle. The particle corresponding to the current global optimum. This refers to the i-th particle in this iteration.

[0097] It is important to note that The difference array is formed by the differences between the parameters of the particle corresponding to the local optimum and the corresponding parameters of the i-th particle. This array reflects the difference between the i-th particle and the particle corresponding to the local optimum in the current iteration. Similarly, This reflects the difference between the i-th particle in the current iteration and the particle corresponding to the global optimum. To ensure that each particle converges towards both the particle corresponding to the local optimum and the particle corresponding to the global optimum, when the i-th particle has a large difference from both the local and global optimum, it is highly likely that the update difference in the next iteration will be larger, accelerating the optimization speed of that particle. Conversely, when the i-th particle has a small difference from both the local and global optimum, it is highly likely that the update difference in the next iteration will be relatively small, because that particle is already relatively superior. The above formula is used to achieve the individual update iterations of each particle.

[0098] in, The difference between each configuration parameter in the particle corresponding to the local optimum and the corresponding configuration parameter in the i-th particle is obtained.

[0099] The configuration parameter of each particle corresponding to the global optimal value is subtracted from the corresponding configuration parameter of the i-th particle, thereby obtaining the difference between the i-th particle and the particle corresponding to the global optimal value.

[0100] In step 202, each particle in the current iteration is updated according to the update difference corresponding to the particle, thereby obtaining the particle corresponding to the next iteration, and further obtaining the input particle swarm of the next iteration.

[0101] For the update of each particle, the corresponding expression is:

[0102] ;

[0103] wherein, is the i-th particle in the n+1-th iteration, is the i-th particle in the n-th iteration, is the update difference of the i-th particle in the n+1-th iteration.

[0104] Further, in the embodiment, the termination condition involves the following design, as shown in the following formula: Figure 5 The method flow includes:

[0105] In step 301, a preset iteration number is set.

[0106] In step 302, it is judged whether the current iteration round is greater than or equal to the preset iteration number.

[0107] In step 303, when the current iteration round is greater than or equal to the preset iteration number, the termination condition is met.

[0108] In step 304, when the current iteration round is less than the preset iteration number, the termination condition is not met.

[0109] Preferably, the formula of the initial update difference of the particle in the initial particle swarm is:

[0110] ;

[0111] wherein, is the initial update difference of the i-th particle in the initial particle swarm, is the particle in which all parameters in the particle are maximum, is the particle in which all parameters in the particle are minimum, is the preset iteration number.

[0112] As shown in the following formula: Figure 6As shown is a device schematic diagram of the serdes chip performance control parameter optimization device according to an embodiment of the present application. The serdes chip performance control parameter optimization device according to the embodiment includes one or more processors 41 and a memory 42.

[0113] The processor 41 and the memory 42 can be connected through a bus or other means, Figure 6 For example, the connection through the bus is taken as an example.

[0114] The memory 42, as a non-volatile computer readable storage medium, can be used to store non-volatile software programs and non-volatile computer executable programs, such as the serdes chip performance control parameter optimization method in the above embodiment. The processor 41 executes the non-volatile software programs and instructions stored in the memory 42, thereby executing the serdes chip performance control parameter optimization method.

[0115] The memory 42 can include a high-speed random access memory, and can also include a non-volatile memory, for example, at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 42 can optionally include a memory remotely arranged with respect to the processor 41, and these remote memories can be connected to the processor 41 through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0116] The program instructions / modules are stored in the memory 42, and when executed by the one or more processors 41, the serdes chip performance control parameter optimization method in the above embodiment is executed, for example, the above-described Figures 1 to 5 Each step shown.

[0117] The embodiment of the present application also provides a computer storage medium, and the computer storage medium stores computer program instructions; the computer program instructions are executed by a processor to implement the serdes chip performance control parameter optimization method provided by the embodiment of the present application.

[0118] The above only describes the preferred embodiments of the present application and should not be used to limit the present application, and any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A serdes chip performance control parameter optimization method, characterized in that, The method comprises the following steps: setting a value range of all configuration parameters of a preset configuration type in a serdes chip, and generating an initial particle swarm according to each configuration parameter and a corresponding value range; inputting the initial particle swarm as an input particle swarm into a first iteration round; each iteration round comprises the following steps: inputting each particle in the input particle swarm into a corresponding serdes chip, obtaining an eye diagram corresponding to each particle, and obtaining a target value corresponding to each particle according to the eye diagram corresponding to each particle; obtaining a maximum target value of each particle in a historical iteration round as a local optimal value of the particle, and obtaining a maximum target value of all particles in all iteration rounds as a global optimal value; and determining whether a termination condition is met; when the termination condition is met, configuration parameters in a particle corresponding to the current global optimal value are taken as optimal configurations; when the termination condition is not met, each particle in the current input particle swarm is updated according to the local optimal value and the global optimal value of the particle, an input particle swarm of a next iteration round is obtained and inputted into the next iteration round; each particle in the current iteration round obtains an update difference corresponding to the particle according to the local optimal value and the global optimal value of the particle; each particle in the current iteration round is updated according to the update difference corresponding to the particle, a particle corresponding to the next iteration round is obtained, and then the input particle swarm of the next iteration round is obtained; and a formula of the update difference is: ； wherein, is the particle update difference in the n-th iteration, is the particle update difference in the n+1-th iteration, and is a random number with a value range of is a random number with a value range of is the particle corresponding to the local optimal value in the current iteration, is the particle corresponding to the current global optimal value, is the i-th particle in the current iteration.​​ 2. The method for serdes chip performance control parameter optimization according to claim 1, characterized in that, the target value corresponding to each particle is obtained according to the eye diagram corresponding to each particle, which comprises the following steps: ； wherein is the target value, is the eye height in the eye diagram, is the weight of the eye height, is the eye width in the eye diagram, is the weight of the eye width.

3. The method for serdes chip performance control parameter optimization according to claim 1, characterized in that, the preset configuration type comprises one or more of a continuous-time linear equalizer parameter, an amplifier parameter, a tap coefficient of a feedback decision equalizer, a proportional path parameter, an integral path parameter, and a tap coefficient of a discrete-time linear equalizer.

4. The method for serdes chip performance control parameter optimization according to claim 3, characterized in that, the value range of all configuration parameters of the preset configuration type in the serdes chip is set, which comprises the following steps: The continuous time linear equalizer parameter has a value range of: wherein is a parameter bit width of the continuous time linear equalizer parameter. The amplifier parameter has a value range of: wherein is a parameter bit width of a continuous-time linear equalizer parameter; The range of the value of each tap coefficient of the feedback decision equalizer is: wherein is the bit width of the i-th tap coefficient of the feedback decision equalizer. The value range of the proportional path parameter is: wherein is a parameter bit width of the proportional path parameter; The integral path parameter has a value range of: wherein is a parameter bit width of the integral path parameter; The range of values for each tap coefficient of the discrete-time linear equalizer is: wherein is the bit width of the i-th tap coefficient of the discrete-time linear equalizer.

5. The method for serdes chip performance control parameter optimization according to claim 1, characterized in that, the termination condition comprises the following steps: a preset iteration number is set; whether a current iteration round is greater than or equal to the preset iteration number is determined; when the current iteration round is greater than or equal to the preset iteration number, the termination condition is met; when the current iteration round is less than the preset iteration number, the termination condition is not met.

6. The method for serdes chip performance control parameter optimization according to claim 5, characterized in that, a formula of an initial update difference of a particle in the initial particle swarm is: ； wherein, is the initial update difference of the i-th particle in the initial particle group, is the particle with all parameters being maximum in the particle, is the particle with all parameters being minimum in the particle, is the preset iteration number.

7. A serdes chip performance control parameter optimization device, characterized in that, The method comprises the following steps:

8. A non-transitory computer storage medium, comprising, The computer storage medium stores computer program instructions, and the computer program instructions are executed by one or more processors to implement the serdes chip performance control parameter optimization method in any one of claims 1 to 6.

Images