CN108122598B - Soft error rate prediction method and system of SRAM with EDAC function - Google Patents

Soft error rate prediction method and system of SRAM with EDAC function Download PDF

Info

Publication number
CN108122598B
CN108122598B CN201711368165.4A CN201711368165A CN108122598B CN 108122598 B CN108122598 B CN 108122598B CN 201711368165 A CN201711368165 A CN 201711368165A CN 108122598 B CN108122598 B CN 108122598B
Authority
CN
China
Prior art keywords
error rate
soft error
edac
sram device
function
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201711368165.4A
Other languages
Chinese (zh)
Other versions
CN108122598A (en
Inventor
张战刚
雷志锋
李沙金
黄云
恩云飞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Electronic Product Reliability and Environmental Testing Research Institute
Original Assignee
China Electronic Product Reliability and Environmental Testing Research Institute
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Electronic Product Reliability and Environmental Testing Research Institute filed Critical China Electronic Product Reliability and Environmental Testing Research Institute
Priority to CN201711368165.4A priority Critical patent/CN108122598B/en
Publication of CN108122598A publication Critical patent/CN108122598A/en
Application granted granted Critical
Publication of CN108122598B publication Critical patent/CN108122598B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G11—INFORMATION STORAGE
    • G11C—STATIC STORES
    • G11C29/00—Checking stores for correct operation ; Subsequent repair; Testing stores during standby or offline operation
    • G11C29/56—External testing equipment for static stores, e.g. automatic test equipment [ATE]; Interfaces therefor
    • G11C29/56008—Error analysis, representation of errors

Landscapes

  • Semiconductor Memories (AREA)
  • Tests Of Electronic Circuits (AREA)

Abstract

The invention relates to the field of radiation effect of electronic devices, in particular to a soft error rate prediction method and a soft error rate prediction system of an SRAM with an EDAC function, which are characterized in that the original soft error rate of the SRAM device with the EDAC function is obtained; acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function; and acquiring the target soft error rate according to the original soft error rate and the incidence relation. In the scheme, the target soft error rate of the SRAM device with the EDAC function can be obtained according to the original soft error rate of the SRAM device and the relation between the target soft error rate and the original soft error rate, and the accuracy of the predicted result of the space soft error rate of the SRAM device with the EDAC function can be improved.

Description

Soft error rate prediction method and system of SRAM with EDAC function
Technical Field
The invention relates to the field of radiation effect of electronic devices, in particular to a soft error rate prediction method and a soft error rate prediction system of an SRAM with an EDAC function.
Background
The universe environment of the earth is full of various Galaxy universe rays, solar universe rays and geomagnetic fields to capture high-energy protons, alpha particles and heavy ions in the band, so that various electronic devices are easy to generate single-particle effect, and various faults of the electronic devices are caused. Among the multiple single event effect types, the proportion of the number of faults caused by single event upset is the largest.
The single event upset refers to the fact that high-energy charged particles enter a sensitive area of a semiconductor device, so that the logic state of the device is inverted to the opposite state, and stored information is in error, and the error is called as a spatial soft error. In order to improve the processing capability of the electronic device for dealing with the single event effect, the spatial soft error rate needs to be effectively analyzed.
In the traditional technology, an RPP/IRPP (Rectangular paralleliped/Integral Rectangular paralleliped, cube/Integral cube) model is adopted to predict the spatial single event effect, only a single sensitive volume is considered in the model, and soft errors can be caused as long as the charge collection amount in the single sensitive volume exceeds critical charge and the single event effect occurs; for an SRAM (Static Random Access Memory) device with an Error Detection and Correction (EDAC) function, even if a single particle effect occurs in a single sensitive volume, due to the adoption of EDAC reinforcement measures, errors caused by the single sensitive volume can be corrected, and soft errors can be caused only when a plurality of errors occur in a single word in a single refresh period, that is, a plurality of sensitive volumes need to be considered. Therefore, when the traditional RPP/IRPP model is used for obtaining the space soft error rate of the SRAM device with the EDAC function, the obtained result has lower accuracy.
Disclosure of Invention
Therefore, it is necessary to provide a method and a system for predicting the soft error rate of the SRAM with EDAC function, aiming at the problem that the accuracy of the obtained result is low when the conventional RPP/IRPP model is used to obtain the spatial soft error rate of the SRAM with EDAC function.
A soft error rate prediction method of an SRAM with an EDAC function comprises the following steps:
acquiring an original soft error rate of an SRAM device with an EDAC function, wherein the original soft error rate is the soft error rate of the SRAM device with the EDAC function under a space application condition when the EDAC function is closed;
acquiring the incidence relation between the target soft error rate of the SRAM device with the EDAC function and the original soft error rate, wherein the target soft error rate is the soft error rate of the SRAM device with the EDAC function under the space application condition when the EDAC function is started;
and acquiring the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate and the incidence relation.
In one embodiment, the step of obtaining the raw soft error rate of the EDAC capable SRAM device comprises the steps of:
closing the EDAC function of the SRAM device with the EDAC function;
irradiating the SRAM device with the closed EDAC function by using a ground accelerator to obtain a single-particle upset section parameter of the SRAM device with the EDAC function;
acquiring particle parameters of particles used in an irradiation process;
and acquiring the original soft error rate according to the particle parameters and the single-particle upset section parameters.
In one embodiment, the step of obtaining the correlation between the target soft error rate and the original soft error rate of the EDAC-enabled SRAM device comprises the steps of:
obtaining a refresh period of an SRAM device, a bit number in a single word of the SRAM device and a sensitive bit number of an EDAC circuit;
acquiring the probability that the number of bits of the SRAM device with errors in a single word in a single refresh period is 2 according to the original soft error rate, the refresh period of the SRAM device and the number of bits in the single word of the SRAM device;
acquiring the error probability of the EDAC circuit according to the original soft error rate and the sensitive digit of the EDAC circuit;
and acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function according to the probability that the number of bits of the error in a single word of the SRAM device is 2 in a single refreshing period and the probability that the error occurs in the EDAC circuit.
In one embodiment, the probability that an error occurs in a word in a single refresh cycle of the SRAM device with a bit number of 2 is obtained according to the following functional relationship:
Figure BDA0001513288670000031
wherein R isSRAMIs SRThe probability of an AM device having a 2 bit error in a single word in a single refresh cycle, TscrubFor refresh period of SRAM device, NbBeing the number of bits within a single word, RrawIs the original soft error rate.
In one embodiment, the probability of error occurrence of the EDAC circuit is obtained according to the following functional relationship:
REDAC＝NEDAC×Rraw
wherein R isEDACProbability of error for EDAC circuit, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate.
In one embodiment, the step of obtaining the correlation between the target soft error rate and the original soft error rate of the EDAC-enabled SRAM device comprises the steps of:
acquiring a plurality of groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function through a ground accelerator;
and fitting the plurality of groups of data points, and acquiring the association relation according to a fitting result.
In one embodiment, the step of obtaining the correlation between the target soft error rate and the original soft error rate of the EDAC-enabled SRAM device comprises the steps of:
acquiring a plurality of groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function through a ground accelerator;
performing optimal fitting on the plurality of groups of data points according to the following functional relation to obtain the correction quantity of the functional relation:
Figure BDA0001513288670000032
wherein R issystemTo target Soft error Rate, TscrubFor refresh period of SRAM device, NwNumber of words, N, for SRAM devicesbNumber of bits in a single word, NEDACBeing EDAC circuitsNumber of sensitive bits, RrawIs the original soft error rate;
and acquiring the incidence relation between the target soft error rate and the original soft error rate according to the correction quantity of the functional relation and the functional relation.
In one embodiment, the step of obtaining a plurality of data points of an experimental raw soft error rate and an experimental target soft error rate of the EDAC-enabled SRAM device by a ground accelerator comprises the steps of:
and irradiating the SRAM device with the EDAC function for multiple times by a ground accelerator, changing the type of used charged particles in each irradiation process, and acquiring multiple groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function.
In one embodiment, the step of obtaining a plurality of data points of an experimental raw soft error rate and an experimental target soft error rate of the EDAC-enabled SRAM device by a ground accelerator comprises the steps of:
and irradiating the SRAM device with the EDAC function for multiple times by a ground accelerator, changing the beam fluence rate of used charged particles in each irradiation process, and acquiring multiple groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function.
According to the soft error rate prediction method of the SRAM with the EDAC function, the original soft error rate of the SRAM device with the EDAC function is obtained; acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function; and acquiring the target soft error rate according to the original soft error rate and the incidence relation. In the scheme, the target soft error rate of the SRAM device with the EDAC function can be obtained according to the original soft error rate of the SRAM device and the relation between the target soft error rate and the original soft error rate.
A soft error rate prediction system for EDAC enabled SRAM comprising the following modules:
the device comprises an original soft error rate acquisition module, a soft error rate calculation module and a soft error rate calculation module, wherein the original soft error rate acquisition module is used for acquiring the original soft error rate of the SRAM device with the EDAC function, and the original soft error rate is the soft error rate of the SRAM device with the EDAC function under the space application condition when the EDAC function is closed;
a soft error rate relationship analysis module, configured to obtain an association relationship between a target soft error rate of the SRAM device with EDAC function and the original soft error rate, where the target soft error rate is a soft error rate of the SRAM device with EDAC function under a space application condition when starting the EDAC function;
and the target soft error rate acquisition module is used for acquiring the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate and the incidence relation.
According to the soft error rate prediction system of the SRAM with the EDAC function, the original soft error rate obtaining module obtains the original soft error rate of the SRAM device with the EDAC function; a soft error rate relation analysis module acquires the incidence relation between the target soft error rate of the SRAM device with the EDAC function and the original soft error rate; and the target soft error rate acquisition module acquires the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate and the incidence relation. In the scheme, the system can obtain the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate of the SRAM device and the relation between the target soft error rate and the original soft error rate.
Drawings
FIG. 1 is a schematic flow chart illustrating a method for predicting soft error rate of an SRAM with EDAC function according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a process for obtaining raw soft error rate of an SRAM device with EDAC functionality according to an embodiment of the present invention;
FIG. 3 is a schematic flow chart illustrating the process of obtaining the correlation between the target soft error rate and the original soft error rate of the SRAM device with EDAC function according to an embodiment of the present invention;
FIG. 4 is a schematic flow chart illustrating the process of obtaining the correlation between the target soft error rate and the original soft error rate of the SRAM device with EDAC function according to an embodiment of the present invention;
FIG. 5 is a schematic flow chart illustrating the process of obtaining the correlation between the target soft error rate and the original soft error rate of the SRAM device with EDAC function according to an embodiment of the present invention;
FIG. 6 is a schematic structural diagram of a soft error rate estimation system of an SRAM with EDAC function according to an embodiment of the present invention;
FIG. 7 is a block diagram of the original soft error rate obtaining module 210 according to an embodiment of the present invention;
FIG. 8 is a block diagram of soft error rate relationship analysis module 220 according to an embodiment of the present invention;
FIG. 9 is a block diagram of soft error rate relationship analysis module 220 according to an embodiment of the present invention;
FIG. 10 is a block diagram of soft error rate relationship analysis module 220 according to an embodiment of the present invention;
FIG. 11 is a flow chart illustrating a method for predicting soft error rate of an SRAM with EDAC function according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the detailed description and specific examples, while indicating the scope of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
Referring to fig. 1, a schematic flow chart of an embodiment of a soft error rate estimation method of an SRAM with EDAC function according to the present invention is shown. The soft error rate prediction method of the SRAM with EDAC function in this embodiment includes the steps of:
step S110: acquiring an original soft error rate of an SRAM device with an EDAC function, wherein the original soft error rate is the soft error rate of the SRAM device with the EDAC function under a space application condition when the EDAC function is closed;
in this step, the space application condition includes an actual application place of the SRAM device with EDAC function; in order to obtain the correlation between the original soft error rate and the target soft error rate, the soft error rate of the SRAM device when the EDAC function is not turned on needs to be obtained.
In another embodiment, referring to fig. 2, the step of obtaining the raw soft error rate of the EDAC-enabled SRAM device in step S110 includes the following steps:
step S111: closing the EDAC function of the SRAM device with the EDAC function;
step S112: irradiating the SRAM device with the closed EDAC function by using a ground accelerator to obtain a single-particle upset section parameter of the SRAM device with the EDAC function;
step S113: acquiring particle parameters of particles used in an irradiation process;
step S114: and acquiring the original soft error rate according to the particle parameters and the single-particle upset section parameters.
In the embodiment, the ground accelerator accelerates specific particles to irradiate the SRAM device, so that a single event effect caused when high-energy particles in a space environment pass through the SRAM device can be simulated, flight tests or experiments in a universe environment are not required, and the process of acquiring the original soft error rate is simplified.
Specifically, the step S112 may adopt a "detect motherboard-load daughter board" separation structure in one embodiment, and obtain the original soft error rate through the ground accelerator. The detection mainboard can be provided with a power supply management module, a temperature monitoring module, a current monitoring module, a device to be detected monitoring module and a communication module; the bearing daughter board can provide different connection interfaces according to the packaging modes of different SRAM devices; and no semiconductor device is arranged between the main board and the bearing daughter board so as to ensure that the generated single event effect is totally contributed by the SRAM device. Irradiating the SRAM device on the bearing daughter board by using a ground accelerator, and continuously performing read-write operation on the SRAM through the detection main board to obtain the position of the bit with the error, thereby determining the single-particle upset section parameter.
Optionally, in step S112, according to the type of the high-energy particles in the actual space environment, selecting specific particles to irradiate through the ground accelerator; for example, if the ground accelerator selects protons for irradiation, the particle parameter in step S113 may be proton energy; if the ground accelerator selects heavy ions for irradiation, the particle parameter in step S113 may be an LET (Linear Energy transfer) value of the heavy ions.
Step S120: acquiring the incidence relation between the target soft error rate of the SRAM device with the EDAC function and the original soft error rate, wherein the target soft error rate is the soft error rate of the SRAM device with the EDAC function under the space application condition when the EDAC function is started;
in another embodiment, referring to fig. 3, the step of obtaining the correlation between the target soft error rate and the original soft error rate of the EDAC-enabled SRAM device in step S120 includes the following steps:
step S121: obtaining a refresh period of an SRAM device, a bit number in a single word of the SRAM device and a sensitive bit number of an EDAC circuit;
step S122: acquiring the probability that the number of bits of the SRAM device with errors in a single word in a single refresh period is 2 according to the original soft error rate, the refresh period of the SRAM device and the number of bits in the single word of the SRAM device;
step S123: acquiring the error probability of the EDAC circuit according to the original soft error rate and the sensitive digit of the EDAC circuit;
step S124: and acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function according to the probability that the number of bits of the error in a single word of the SRAM device is 2 in a single refreshing period and the probability that the error occurs in the EDAC circuit.
In this embodiment, considering that a hamming code is selected as an ecc (error Correcting code) code by a common EDAC circuit, a 1-bit error occurring in an SRAM device can be corrected, a soft error may be caused only when the number of bits of the error occurring in the SRAM device with the EDAC function exceeds 1 bit, and a general SRAM device has a refresh mechanism, and the bit where the error occurs can be corrected by regularly updating stored data; in addition, the EDAC circuit is used as a reinforcing measure of the SRAM device, once the EDAC circuit generates errors, the soft error of the SRAM device can be misjudged, so that the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function can be more accurately acquired according to the probability that the number of bits of the errors generated in a single word of the SRAM device in a single refresh period is a preset value 2 and the probability that the EDAC circuit generates the errors; the refreshing period is the interval time of updating the stored data of the SRAM device with the EDAC function; the radiation sensitive characteristic parameters of the EDAC circuit comprise the bit number and the position of the sensitive bit of the EDAC circuit, and the sensitive bit of the EDAC circuit is a bit which can cause errors after logic inversion occurs.
Specifically, in one embodiment, the sensitive characteristic parameters of the EDAC circuit may be obtained by a method of bit-by-bit flipping fault injection: and carrying out bit-by-bit overturning on the EDAC circuit, comparing the overturning detection with the expected output after each time of overturning detection, judging that a bit is a sensitive bit if the actual output result caused by the overturning of a certain bit is different from the expected output result, and recording the number and the position of the sensitive bit.
In another embodiment, the probability that the number of bits of an error in a single word in a single refresh cycle of the SRAM device in step S122 is 2 is obtained according to the following functional relationship:
Figure BDA0001513288670000081
wherein R isSRAMIs the probability of the SRAM device having a number of 2 bits of error in a single word in a single refresh cycle, TscrubFor refresh period of SRAM device, NbBeing the number of bits within a single word, RrawIs the original soft error rate.
In this embodiment, RSRAMProbability of 2 bits of error occurring in a single word in a single refresh cycle for an SRAM device, whereA single word median comprising data bits and check bits; because the error can be corrected by the EDAC circuit when the number of bits of the error occurring in the SRAM device with the EDAC function is less than or equal to 1 bit, and the SRAM device can update the stored data through the refresh mechanism, the probability that the number of bits of the error occurring in a single word of the SRAM device in a single refresh period is 2 can be obtained through the functional relational expression of this embodiment, so that the correlation between the target soft error rate and the original soft error rate is obtained more accurately.
In another embodiment, the error probability of the EDAC circuit in step S123 is obtained according to the following functional relationship:
REDAC＝NEDAC×Rraw
wherein R isEDACProbability of error for EDAC circuit, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate.
In this embodiment, the sensitive bit of the EDAC circuit refers to a bit that causes an output error after a flip occurs in the EDAC circuit; the probability of error occurrence of the EDAC circuit can be accurately obtained according to the functional relation of the embodiment.
In another embodiment, the step S124 obtains the correlation between the target soft error rate of the SRAM device with EDAC function and the original soft error rate, which is the following functional relation:
Rsystem＝RSRAM+REDAC
wherein R issystemTo target Soft error Rate, RSRAMFor SRAM devices the probability of an error occurring in a single word in a single refresh cycle is 2, REDACThe probability of error for the EDAC circuit.
In this embodiment, the target soft error rate of the SRAM device with EDAC function is related to the probability of the number of error bits occurring in a single word being 2 in a single refresh cycle, and also related to the radiation sensitivity of the EDAC circuit.
In another embodiment, referring to fig. 4, the step of obtaining the correlation between the target soft error rate and the original soft error rate of the EDAC-enabled SRAM device in step S120 includes the following steps:
step S124: acquiring a plurality of groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function through a ground accelerator;
step S125: and fitting the plurality of groups of data points, and acquiring the association relation according to a fitting result.
In this embodiment, the experimental target soft error rate and the experimental original soft error rate are respectively a soft error rate when the EDAC function of the SRAM device with EDAC function is turned on and a soft error rate when the EDAC function is not turned on under the experimental condition, and the irradiation of the SRAM device with EDAC function by the ground accelerator can simulate the radiation environment under the spatial application condition, and obtain a plurality of groups of data points related to the experimental target soft error rate and the experimental original soft error rate; the incidence relation of the SRAM device with the EDAC function under the experimental condition can be obtained by fitting the obtained multiple groups of data points, and the incidence relation is suitable for a space application environment.
In another embodiment, referring to fig. 5, the step of obtaining the correlation between the target soft error rate and the original soft error rate of the EDAC-enabled SRAM device in step S120 includes the following steps:
step S124: acquiring a plurality of groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function through a ground accelerator;
step S126: performing optimal fitting on the plurality of groups of data points according to the following functional relation to obtain the correction quantity of the functional relation:
Figure BDA0001513288670000101
wherein R issystemTo target Soft error Rate, TscrubIs SRARefresh period of M devices, NwNumber of words, N, for SRAM devicesbNumber of bits in a single word, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate;
step S127: and acquiring the incidence relation between the target soft error rate and the original soft error rate according to the correction quantity of the functional relation and the functional relation.
In this embodiment, the multiple groups of data points obtained by the ground accelerator are experimental measurement data, which include an experimental original soft error rate and an experimental target soft error rate, the functional relation is a theoretical relationship obtained according to specific parameters of the SRAM with EDAC function, the multiple groups of data points are optimally fitted according to the functional relation, the theoretical relationship can be corrected by using the experimental measurement data, the correlation obtained after correction is more accurate, and the obtained correlation is suitable for prediction of the target soft error rate of the SRAM with EDAC function in a spatial environment.
In another embodiment, the step of obtaining the plurality of sets of data points of the original soft error rate and the target soft error rate of the EDAC-enabled SRAM device by the ground accelerator in step S124 includes the steps of:
and irradiating the SRAM device with the EDAC function for multiple times by a ground accelerator, changing the type of used charged particles in each irradiation process, and acquiring multiple groups of data points of the original soft error rate and the target soft error rate of the SRAM device with the EDAC function.
In this embodiment, since the space application environment includes a plurality of types of high-energy particles, and the degree of soft errors of the SRAM device caused by different high-energy particles is different, in the process of irradiating the SRAM device with EDAC function by using the ground accelerator, the irradiation environment for space application can be simulated by changing the type of charged particles used for irradiation, so that the obtained plurality of sets of data points are closer to reality.
In another embodiment, the step of obtaining the plurality of sets of data points of the original soft error rate and the target soft error rate of the EDAC-enabled SRAM device by the ground accelerator in step S124 includes the steps of:
and irradiating the SRAM device with the EDAC function for multiple times by a ground accelerator, changing the beam fluence rate of used charged particles in each irradiation process, and acquiring multiple groups of data points of the original soft error rate and the target soft error rate of the SRAM device with the EDAC function.
In this embodiment, because the energy carried by the high-energy particles each time the high-energy particles pass through the SRAM device is different under the space application condition, and the degrees of soft errors of the SRAM device caused by the high-energy particles carrying different energies are different, in the process of irradiating the SRAM device with the EDAC function by using the ground accelerator, the irradiation environment for the space application can be simulated by changing the beam fluence rate of the charged particles used for irradiation, so that the obtained multiple groups of data points are closer to reality.
Step S130: and acquiring the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate and the incidence relation.
According to the soft error rate prediction method of the SRAM with the EDAC function, the original soft error rate of the SRAM device with the EDAC function is obtained; acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function; and acquiring the target soft error rate according to the original soft error rate and the incidence relation. In the scheme, the target soft error rate of the SRAM device with the EDAC function can be obtained according to the original soft error rate of the SRAM device and the relation between the target soft error rate and the original soft error rate.
Fig. 6 is a schematic structural diagram of a soft error rate prediction system of an SRAM with EDAC function according to an embodiment of the present invention. The soft error rate prediction system of the EDAC capable SRAM in this embodiment includes the following modules:
an original soft error rate obtaining module 210, configured to obtain an original soft error rate of the SRAM device with EDAC function, where the original soft error rate is a soft error rate of the SRAM device with EDAC function under a space application condition when the EDAC function is turned off;
in another embodiment, referring to fig. 7, the raw soft error rate acquisition module 210 comprises the following modules:
a function shutdown module 211, configured to shut down an EDAC function of the SRAM device with the EDAC function;
the accelerated irradiation module 212 is configured to irradiate the SRAM device with the EDAC function turned off by using a ground accelerator, and obtain a single-particle upset section parameter of the SRAM device with the EDAC function;
a particle parameter obtaining module 213, configured to obtain particle parameters of particles used in the irradiation process;
an error rate obtaining module 214, configured to obtain the original soft error rate according to the particle parameter and the single-particle upset cross-section parameter.
A soft error rate relationship analysis module 220, configured to obtain an association relationship between a target soft error rate of the SRAM device with EDAC function and the original soft error rate, where the target soft error rate is a soft error rate of the SRAM device with EDAC function under a space application condition when starting the EDAC function;
in another embodiment, referring to FIG. 8, the soft error rate relationship analysis module 220 includes the following modules:
the parameter obtaining module 221 is configured to obtain a refresh period of the SRAM device, a bit number in a single word of the SRAM device, and a sensitive bit number of the EDAC circuit;
an SRAM error rate obtaining module 222, configured to obtain, according to the original soft error rate, a refresh period of the SRAM device, and a bit number in a single word of the SRAM device, a probability that a bit number of an error occurring in the single word of the SRAM device in the single refresh period is 2;
the EDAC error rate obtaining module 223 is used for obtaining the error probability of the EDAC circuit according to the original soft error rate and the sensitive digit of the EDAC circuit;
and the incidence relation obtaining module 224 is configured to obtain the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function according to the probability that the number of bits of the error in a single word of the SRAM device in a single refresh cycle is 2 and the probability that the error occurs in the EDAC circuit.
In another embodiment, the SRAM error rate obtaining module 222 obtains the probability that the SRAM device has an error of 2 bits in a single word in a single refresh cycle according to the following functional relationship:
Figure BDA0001513288670000131
wherein R isSRAMIs the probability of the SRAM device having a number of 2 bits of error in a single word in a single refresh cycle, TscrubFor refresh period of SRAM device, NbBeing the number of bits within a single word, RrawIs the original soft error rate.
In another embodiment, the EDAC error rate obtaining module 223 obtains the error probability of the EDAC circuit according to the following functional relation:
REDAC＝NEDAC×Rraw
wherein R isEDACProbability of error for EDAC circuit, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate.
In another embodiment, the association relationship obtained by the association relationship obtaining module 224 is the following functional relationship:
Rsystem＝RSRAM+REDAC
wherein R issystemTo target Soft error Rate, RSRAMFor SRAM devices the probability of an error occurring in a single word in a single refresh cycle is 2, REDACThe probability of error for the EDAC circuit.
In another embodiment, referring to FIG. 9, the soft error rate relationship analysis module 220 includes the following modules:
a data point obtaining module 224, configured to obtain, through a ground accelerator, multiple sets of data points of an experimental original soft error rate and an experimental target soft error rate of the SRAM device with EDAC function;
and a data point fitting module 225, configured to fit the multiple sets of data points, and obtain the association relationship according to a fitting result.
In another embodiment, referring to FIG. 10, the soft error rate relationship analysis module 220 includes the following modules:
a data point obtaining module 224, configured to obtain, through a ground accelerator, multiple sets of data points of an experimental original soft error rate and an experimental target soft error rate of the SRAM device with EDAC function;
a correction amount obtaining module 226, configured to perform optimal fitting on the multiple sets of data points according to the following functional relation, so as to obtain a correction amount of the functional relation:
Figure BDA0001513288670000141
wherein R issystemTo target Soft error Rate, TscrubFor refresh period of SRAM device, NwNumber of words, N, for SRAM devicesbNumber of bits in a single word, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate;
and the relation correction module 227 is configured to obtain an association relation between the target soft error rate and the original soft error rate according to the correction amount of the functional relation and the functional relation.
In another embodiment, the data point obtaining module 224 irradiates the SRAM device with EDAC function through the ground accelerator multiple times, and changes the kind of the used charged particles during each irradiation process to obtain multiple sets of data points of the original soft error rate and the target soft error rate of the SRAM device with EDAC function.
In another embodiment, the data point obtaining module 224 irradiates the SRAM device with EDAC function through a ground accelerator for multiple times, changes a beam fluence rate of charged particles used in each irradiation process, and obtains multiple sets of data points of an original soft error rate and a target soft error rate of the SRAM device with EDAC function.
And a target soft error rate obtaining module 230, configured to obtain a target soft error rate of the SRAM device with EDAC function according to the original soft error rate and the association relationship.
The soft error rate prediction system of the SRAM with the EDAC function and the soft error rate prediction method of the SRAM with the EDAC function correspond to each other, and the technical characteristics and the beneficial effects described in the embodiment of the soft error rate prediction method of the SRAM with the EDAC function are applicable to the embodiment of the soft error rate prediction system of the SRAM with the EDAC function.
Referring to fig. 11, a schematic flow chart of an embodiment of a soft error rate estimation method of an SRAM with EDAC function according to the present invention is shown. In this embodiment, the ECC code used by the EDAC circuit of the SRAM device with EDAC function is a hamming code, and the soft error rate prediction method of the SRAM with EDAC function includes the following steps:
step S311: irradiating the SRAM device with the closed EDAC function by using a ground accelerator to obtain a single-particle upset section parameter of the SRAM device with the EDAC function;
step S312: acquiring particle parameters of particles used in an irradiation process;
step S313: acquiring an original soft error rate according to the particle parameters and the single-particle upset section parameters;
step S321: irradiating the SRAM device with the EDAC function for multiple times through a ground accelerator, changing the type of used charged particles or beam fluence rate in each irradiation process, and acquiring multiple groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function;
step S322: performing optimal fitting on the plurality of groups of data points according to the following functional relation to obtain the correction quantity of the functional relation:
Figure BDA0001513288670000151
wherein R issystemTo target Soft error Rate, TscrubFor refresh period of SRAM device, NwNumber of words, N, for SRAM devicesbNumber of bits in a single word, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate;
step S323: acquiring the incidence relation between the target soft error rate and the original soft error rate according to the correction quantity of the functional relation and the functional relation;
step S330: and acquiring the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate and the incidence relation.
According to the soft error rate prediction method of the SRAM with the EDAC function, the original soft error rate of the SRAM device with the EDAC function is obtained; acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function; and acquiring the target soft error rate according to the original soft error rate and the incidence relation. In the scheme, the target soft error rate of the SRAM device with the EDAC function can be obtained according to the original soft error rate of the SRAM device and the relation between the target soft error rate and the original soft error rate.
According to the soft error rate prediction method of the SRAM with the EDAC function, the embodiment of the invention also provides a readable storage medium and a computer device. The readable storage medium stores an executable program, and the program realizes the soft error rate prediction method of the SRAM with the EDAC function when being executed by a processor; the computer device comprises a memory, a processor and an executable program stored on the memory and capable of running on the processor, and the processor executes the program to realize the steps of the soft error rate prediction method of the SRAM with the EDAC function.
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (11)

1. A soft error rate prediction method of an SRAM with an EDAC function is characterized by comprising the following steps:
acquiring an original soft error rate of an SRAM device with an EDAC function, wherein the original soft error rate is the soft error rate of the SRAM device with the EDAC function under a space application condition when the EDAC function is closed;
acquiring the incidence relation between the target soft error rate of the SRAM device with the EDAC function and the original soft error rate, wherein the target soft error rate is the soft error rate of the SRAM device with the EDAC function under the space application condition when the EDAC function is started;
obtaining the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate and the incidence relation, wherein the step of obtaining the incidence relation between the target soft error rate of the SRAM device with the EDAC function and the original soft error rate comprises the following steps: obtaining a refresh period of an SRAM device, a bit number in a single word of the SRAM device and a sensitive bit number of an EDAC circuit;
acquiring the probability that the number of bits of the SRAM device with errors in a single word in a single refresh period is 2 according to the original soft error rate, the refresh period of the SRAM device and the number of bits in the single word of the SRAM device;
acquiring the error probability of the EDAC circuit according to the original soft error rate and the sensitive digit of the EDAC circuit;
and acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function according to the probability that the number of bits of the error in a single word of the SRAM device is 2 in a single refreshing period and the probability that the error occurs in the EDAC circuit.
2. The method of claim 1, wherein the step of obtaining the raw soft error rate of the EDAC capable SRAM device comprises the steps of:
closing the EDAC function of the SRAM device with the EDAC function;
irradiating the SRAM device with the closed EDAC function by using a ground accelerator to obtain a single-particle upset section parameter of the SRAM device with the EDAC function;
acquiring particle parameters of particles used in an irradiation process;
and acquiring the original soft error rate according to the particle parameters and the single-particle upset section parameters.
3. The method of claim 1, wherein the probability of an error occurring in a word of 2 bits in a single refresh cycle of the SRAM device is obtained according to the following functional relationship:
Figure FDA0002508134260000021
wherein R isSRAMIs the probability of the SRAM device having a number of 2 bits of error in a single word in a single refresh cycle, TscrubFor refresh period of SRAM device, NbBeing the number of bits within a single word, RrawIs the original soft error rate.
4. The method of claim 1, wherein the probability of error occurrence for the EDAC circuit is obtained according to the following functional relationship:
REDAC＝NEDAC×RRAW
wherein R isEDACProbability of error for EDAC circuit, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate.
5. A soft error rate prediction method of an SRAM with an EDAC function is characterized by comprising the following steps:
acquiring an original soft error rate of an SRAM device with an EDAC function, wherein the original soft error rate is the soft error rate of the SRAM device with the EDAC function under a space application condition when the EDAC function is closed;
acquiring the incidence relation between the target soft error rate of the SRAM device with the EDAC function and the original soft error rate, wherein the target soft error rate is the soft error rate of the SRAM device with the EDAC function under the space application condition when the EDAC function is started;
obtaining the target soft error rate of the SRAM device with the EDAC function according to the original soft error rate and the incidence relation, wherein the step of obtaining the incidence relation between the target soft error rate of the SRAM device with the EDAC function and the original soft error rate comprises the following steps: acquiring a plurality of groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function through a ground accelerator; and fitting the plurality of groups of data points, and acquiring the association relation according to a fitting result.
6. The method of claim 5, wherein the step of obtaining the raw soft error rate of the EDAC capable SRAM device comprises the steps of:
closing the EDAC function of the SRAM device with the EDAC function;
irradiating the SRAM device with the closed EDAC function by using a ground accelerator to obtain a single-particle upset section parameter of the SRAM device with the EDAC function;
acquiring particle parameters of particles used in an irradiation process;
and acquiring the original soft error rate according to the particle parameters and the single-particle upset section parameters.
7. The method of claim 5, wherein the step of obtaining the correlation between the target soft error rate and the original soft error rate of the EDAC capable SRAM device comprises the steps of:
acquiring a plurality of groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function through a ground accelerator;
performing optimal fitting on the plurality of groups of data points according to the following functional relation to obtain the correction quantity of the functional relation:
Figure FDA0002508134260000031
wherein R issystemTo target Soft error Rate, TscrubFor refresh period of SRAM device, NwNumber of words, N, for SRAM devicesbNumber of bits in a single word, NEDACSensitive bits for EDAC circuits, RrawIs the original soft error rate;
and acquiring the incidence relation between the target soft error rate and the original soft error rate according to the correction quantity of the functional relation and the functional relation.
8. The method of claim 5 or 7, wherein the step of obtaining the plurality of data points of the experimental raw soft error rate and the experimental target soft error rate of the EDAC-capable SRAM device through the ground accelerator comprises the steps of:
and irradiating the SRAM device with the EDAC function for multiple times by a ground accelerator, changing the type of used charged particles in each irradiation process, and acquiring multiple groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function.
9. The method of claim 5 or 7, wherein the step of obtaining the plurality of data points of the experimental raw soft error rate and the experimental target soft error rate of the EDAC-capable SRAM device through the ground accelerator comprises the steps of:
and irradiating the SRAM device with the EDAC function for multiple times by a ground accelerator, changing the beam fluence rate of used charged particles in each irradiation process, and acquiring multiple groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function.
10. A soft error rate prediction system for an SRAM with EDAC capability, comprising the following modules:
the device comprises an original soft error rate acquisition module, a soft error rate calculation module and a soft error rate calculation module, wherein the original soft error rate acquisition module is used for acquiring the original soft error rate of the SRAM device with the EDAC function, and the original soft error rate is the soft error rate of the SRAM device with the EDAC function under the space application condition when the EDAC function is closed;
a soft error rate relationship analysis module, configured to obtain an association relationship between a target soft error rate of the SRAM device with EDAC function and the original soft error rate, where the target soft error rate is a soft error rate of the SRAM device with EDAC function under a space application condition when starting the EDAC function;
a target soft error rate obtaining module, configured to obtain a target soft error rate of the SRAM device with EDAC function according to the original soft error rate and the association relationship, where the step of obtaining the association relationship between the target soft error rate of the SRAM device with EDAC function and the original soft error rate includes the following steps: obtaining a refresh period of an SRAM device, a bit number in a single word of the SRAM device and a sensitive bit number of an EDAC circuit;
acquiring the probability that the number of bits of the SRAM device with errors in a single word in a single refresh period is 2 according to the original soft error rate, the refresh period of the SRAM device and the number of bits in the single word of the SRAM device;
acquiring the error probability of the EDAC circuit according to the original soft error rate and the sensitive digit of the EDAC circuit;
and acquiring the incidence relation between the target soft error rate and the original soft error rate of the SRAM device with the EDAC function according to the probability that the number of bits of the error in a single word of the SRAM device is 2 in a single refreshing period and the probability that the error occurs in the EDAC circuit.
11. A soft error rate prediction system for an SRAM with EDAC capability, comprising the following modules:
the device comprises an original soft error rate acquisition module, a soft error rate calculation module and a soft error rate calculation module, wherein the original soft error rate acquisition module is used for acquiring the original soft error rate of the SRAM device with the EDAC function, and the original soft error rate is the soft error rate of the SRAM device with the EDAC function under the space application condition when the EDAC function is closed;
a soft error rate relationship analysis module, configured to obtain an association relationship between a target soft error rate of the SRAM device with EDAC function and the original soft error rate, where the target soft error rate is a soft error rate of the SRAM device with EDAC function under a space application condition when starting the EDAC function;
a target soft error rate obtaining module, configured to obtain a target soft error rate of the SRAM device with EDAC function according to the original soft error rate and the association relationship, where the step of obtaining the association relationship between the target soft error rate of the SRAM device with EDAC function and the original soft error rate includes the following steps: acquiring a plurality of groups of data points of the experimental original soft error rate and the experimental target soft error rate of the SRAM device with the EDAC function through a ground accelerator; and fitting the plurality of groups of data points, and acquiring the association relation according to a fitting result.
CN201711368165.4A 2017-12-18 2017-12-18 Soft error rate prediction method and system of SRAM with EDAC function Active CN108122598B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711368165.4A CN108122598B (en) 2017-12-18 2017-12-18 Soft error rate prediction method and system of SRAM with EDAC function

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711368165.4A CN108122598B (en) 2017-12-18 2017-12-18 Soft error rate prediction method and system of SRAM with EDAC function

Publications (2)

Publication Number Publication Date
CN108122598A CN108122598A (en) 2018-06-05
CN108122598B true CN108122598B (en) 2020-12-08

Family

ID=62229377

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711368165.4A Active CN108122598B (en) 2017-12-18 2017-12-18 Soft error rate prediction method and system of SRAM with EDAC function

Country Status (1)

Country Link
CN (1) CN108122598B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109669804B (en) * 2018-11-29 2022-04-19 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） Method and apparatus for reducing the actual soft error rate of a storage area of an ECC memory
CN109656746A (en) * 2018-11-29 2019-04-19 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） Method and apparatus for determining the refreshing frequency of ECC memory
CN109669803A (en) * 2018-11-29 2019-04-23 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） For reducing the method and apparatus of the soft error rate of ECC memory
CN112230081B (en) * 2020-09-17 2023-08-25 兰州空间技术物理研究所 Equivalent LET calculation method for pulse laser single event effect test
CN112668932B (en) * 2021-01-15 2023-12-12 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） Reinforcing design method for electronic device, computer equipment and storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102915768A (en) * 2012-10-01 2013-02-06 中国科学院近代物理研究所 Device and method for tolerating faults of storage based on triple modular redundancy of EDAC module
CN103838946A (en) * 2012-11-23 2014-06-04 北京圣涛平试验工程技术研究院有限责任公司 System and method for evaluating protection effectiveness through single particle effect
CN104851467A (en) * 2015-05-20 2015-08-19 中国空间技术研究院 Function verification method for SRAM with EDAC function for use in space navigation
CN104934072A (en) * 2015-05-12 2015-09-23 工业和信息化部电子第五研究所 Single particle effect detection method and system
CN105679370A (en) * 2014-11-21 2016-06-15 北京圣涛平试验工程技术研究院有限责任公司 SRAM neutron single event effect test control method and apparatus

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7840876B2 (en) * 2007-02-20 2010-11-23 Qimonda Ag Power savings for memory with error correction mode
US8510635B2 (en) * 2010-12-28 2013-08-13 United Microelectronics Corp. Method for evaluating failure rate
US8661320B2 (en) * 2011-10-07 2014-02-25 Aeroflex Colorado Springs Inc. Independent orthogonal error correction and detection

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102915768A (en) * 2012-10-01 2013-02-06 中国科学院近代物理研究所 Device and method for tolerating faults of storage based on triple modular redundancy of EDAC module
CN103838946A (en) * 2012-11-23 2014-06-04 北京圣涛平试验工程技术研究院有限责任公司 System and method for evaluating protection effectiveness through single particle effect
CN105679370A (en) * 2014-11-21 2016-06-15 北京圣涛平试验工程技术研究院有限责任公司 SRAM neutron single event effect test control method and apparatus
CN104934072A (en) * 2015-05-12 2015-09-23 工业和信息化部电子第五研究所 Single particle effect detection method and system
CN104851467A (en) * 2015-05-20 2015-08-19 中国空间技术研究院 Function verification method for SRAM with EDAC function for use in space navigation

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
一种改进的SRAM单粒子效应检测系统;童腾 等;《原子核物理评论》;20140630;第31卷(第2期);第170-174页 *
基于故障概率的组合电路软错误率分析;闫爱斌 等;《电子测量与仪器学报》;20150331;第29卷(第3期);第343-346页 *
抗单粒子翻转的双端口SRAM定时刷新机制研究;陈晨 等;《微电子学》;20150831;第45卷(第4期);第512-515页 *

Also Published As

Publication number Publication date
CN108122598A (en) 2018-06-05

Similar Documents

Publication Publication Date Title
CN108122598B (en) Soft error rate prediction method and system of SRAM with EDAC function
Patel et al. Understanding and modeling on-die error correction in modern DRAM: An experimental study using real devices
CN102332310B (en) FPGA (Field Programmable Gate Array)-based single event effect test system for NAND FLASH device
Bautista Gomez et al. Detecting silent data corruption through data dynamic monitoring for scientific applications
US20070011513A1 (en) Selective activation of error mitigation based on bit level error count
CN110988969B (en) Atmospheric neutron radiation effect test method and device
CN108169660B (en) Method and system for distinguishing FPGA single event effect detection data in atmospheric neutron
CN102521467A (en) Bit-by-bit upset fault injection method specifically for SRAM (static random access memory) type FPGA (field programmable gate array)
CN105548866A (en) SRAM type FPGA test method based on irradiation test environment simulation
Luza et al. Emulating the effects of radiation-induced soft-errors for the reliability assessment of neural networks
Wang et al. Neural network based silent error detector
Ponzina et al. E 2 cnns: Ensembles of convolutional neural networks to improve robustness against memory errors in edge-computing devices
US20140203836A1 (en) Method of characterizing the sensitivity of an electronic component subjected to irradiation conditions
Meng et al. A self-test framework for detecting fault-induced accuracy drop in neural network accelerators
CN107144776B (en) A kind of detection method and device of total dose effect
CN107886990A (en) The method of testing and system of the incorgruous single-particle inversion of microprocessor embedded SRAM multidigit
Blume-Kohout et al. Wildcard error: Quantifying unmodeled errors in quantum processors
CN102901924B (en) The method of the single-particle inversion characteristic of a kind of part of detecting triplication redundancy FPGA
CN111985163B (en) Rubidium clock life prediction method and device
CN104700900A (en) Detecting system and method for storage single-event upset effect
CN115421967B (en) Method and system for evaluating storage abnormal risk point of secondary equipment
CN112798944B (en) FPGA hardware error attribution analysis method based on online real-time data
CN112216336B (en) Memory single particle testing method and system
Falguere et al. A statistical method to extract mbu without scrambling information
CN110007738B (en) Method for reconstructing operation state of sensitive circuit after transient ionizing radiation resistance reset

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
CB02 Change of applicant information

Address after: 511300 No.78, west of Zhucun Avenue, Zhucun street, Zengcheng District, Guangzhou City, Guangdong Province

Applicant after: CHINA ELECTRONIC PRODUCT RELIABILITY AND ENVIRONMENTAL TESTING Research Institute

Address before: 510610 No. 110 Zhuang Road, Tianhe District, Guangdong, Guangzhou, Dongguan

Applicant before: CHINA ELECTRONIC PRODUCT RELIABILITY AND ENVIRONMENTAL TESTING Research Institute

CB02 Change of applicant information
GR01 Patent grant
GR01 Patent grant