JP3573669B2 - Semiconductor memory repair operation method and repair operation processing device - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a repair operation method and a repair operation processing device thereof in a process of manufacturing a semiconductor memory. More particularly, the present invention relates to a semiconductor memory capable of executing the repair process at high speed using a small amount of fail information. The present invention relates to a repair operation method and a repair operation processing device.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, repair processing or redundancy operation processing of a semiconductor memory has been known, and various techniques have been disclosed for improving an efficient processing method.
[0003]
For example, in order to increase the capacity of a semiconductor memory, the segment size is increased, the number of segments is increased, and the number of spare lines for performing repair is also increasing. For this reason, a test and repair apparatus using an algorithm as shown in FIG. 15 is disclosed in Japanese Patent Application Laid-Open No. 10-289163, a method for reducing the storage area of fail information, and Japanese Patent Application Laid-Open No. 11-213695. As disclosed in Japanese Patent Application Laid-Open Publication No. H10-209, a method of performing a redundancy operation at a high speed has been proposed.
[0004]
[Problems to be solved by the invention]
However, the prior art has the following problems.
[0005]
The first problem is a decrease in the rescue rate. The reason is that the combination is complicated by the spare line extending over a plurality of segments.
[0006]
In other words, the optimum operation result in one segment may not be the optimum operation result in another segment, and it may be determined that repair is not possible.
[0007]
In a conventional method for determining a line fail first as exemplified in FIG. 15, an example in which repair is possible is shown in FIG. 13, and an example in which repair is impossible is shown in FIG.
[0008]
In the above example, since the X spare lines that can be commonly used for the upper and lower segments have already been determined first, undetermined fail elements remain in the lower segment.
[0009]
The second problem is a delay in processing time. The reason is that it takes time to count and search for fail elements due to an increase in the segment size and spare lines.
[0010]
In the conventional method, since the fail information as shown in FIG. 4 is directly represented in a two-dimensional array, the processing time increases in accordance with the segment size and the spare line.
[0011]
A third problem is an increase in storage area. The reason is that in the conventional means as shown in FIG. 16, the fail information is stored in two dimensions. Therefore, it is necessary to increase the storage area in response to the increase in the number of repairable failures. This is because the proportion of non-element portions increased.
[0012]
In addition, Japanese Patent Application Laid-Open No. 10-107069 discloses a method of determining the order of spare processing using a counter. However, in this method, the means for generating the fail information is the same as the conventional method. Therefore, it does not affect the reduction in the number of memories or the improvement in the repair processing speed. Japanese Patent Application Laid-Open No. H11-16390 describes a method of processing fail information using a tree-structured template. However, in such a method, the repair processing becomes complicated, and there is no effect in improving the processing speed and reducing the storage area.
[0013]
SUMMARY OF THE INVENTION An object of the present invention is to improve the above-described disadvantages of the prior art, divide a fail pattern in a segment into pseudo segments, perform arithmetic processing in consideration of the size thereof, and apply a list-type data configuration. Providing a semiconductor memory repair operation method and a semiconductor memory repair operation apparatus capable of reducing the size of a data storage area by securing only a storage area for fail information to be performed and speeding up arithmetic processing It is.
[0014]
[Means for Solving the Problems]
The present invention employs the following technical configuration to achieve the above object.
[0015]
That is, as a first aspect according to the present invention, a first step of extracting a fail address of a fail element included in a semiconductor memory, and, among the extracted fail addresses,A fail element for extracting a fail element whose one address selected from the X address and the Y address is the same. Two Step, a third step of extracting a fail element having the same address as the fail element extracted in the second step, and a fail element extracted in the third step, wherein the one address is the same as the fail element extracted in the third step. A fourth step of extracting a certain fail element;,A fifth step of repeating the second to fourth steps until there are no more fail elements to be extracted to form a pseudo segment group including the extracted fail elements;Assigning a spare line provided in the semiconductor memory in advance for each pseudo segment group6th stepThe presence or absence of a fail element to which the spare line is not allocated and the presence or absence of an unused spare line.7th stepA second aspect of the present invention relates to a repair operation processing apparatus for a semiconductor memory, which is capable of detecting presence or absence of a failure in each memory from the semiconductor memory. Fail read means for outputting the obtained fail information, and a fail element in which a fail has occurred is extracted from the fail information, and one of the X address and the Y address among the addresses in the semiconductor memory is the same. List creation means for forming a pseudo fail group composed of a fail element, and creating a fail information table including an address of the fail element in the semiconductor element and information indicating which of the pseudo fail groups the fail element belongs to; Storage means for storing the fail information table; In accordance with the stored said fail information table 憶 means, a repair processing apparatus of a semiconductor memory composed of a repair process means for assigning a spare line provided beforehand in the semiconductor memory for each of the pseudo-segment group.
[0016]
BEST MODE FOR CARRYING OUT THE INVENTION
The repair operation method of the semiconductor memory and the repair operation processing device of the semiconductor memory according to the present invention employ the above-described technical configuration. In order to solve the problem, a fail pattern in a segment is divided into pseudo segments, and arithmetic processing is performed in consideration of the size of the segment. In order to solve the third problem, a list-type data structure is applied to secure only a storage area for existing fail information. Therefore, the data storage area size can be reduced, A repair operation method for a semiconductor memory and a repair operation processing device for a semiconductor memory in which the arithmetic processing is also speeded up can be easily obtained.
[0017]
【Example】
Hereinafter, a configuration of a specific example of the repair operation method of the semiconductor memory and the repair operation processing device of the semiconductor memory according to the present invention will be described in detail with reference to the drawings.
[0018]
That is, FIG. 1 shows a repair operation processing apparatus for a semiconductor memory according to the present invention.100 of1 is a block diagram showing a configuration of a specific example, in which a repair processing apparatus 100 in a manufacturing process of a semiconductor memory, in which a fail element in a failed state in a memory to be inspected is replaced with appropriate data. Fail reading means 22 for detecting the presence or absence of a failure for each element from a fail memory 21 in which the result of the failure inspection is detected by using a device, and a fail information output from the fail reading means 22 Only a fail element in which a state has occurred is selected and extracted, and a pseudo segment table 30, a fail information table 31, an X, Y list table 32, and the like are created in accordance with the address of the individual fail element in the memory under test. List creation means 23, and the pseudo-form created by the list creation means 23 And a repair processing unit 24 for executing a repair process in accordance with the information stored in the storage unit 40. 1 shows a repair operation processing device 100 for a semiconductor memory.
[0019]
In other words, the fail memory 21 stores the addresses of the individual fail elements for which the failed state has been confirmed in the memory under test.
[0020]
Further, in FIG. 1, a repair processing result storage means 33 for storing a repair result preferably processed by the repair processing means 24 is provided in addition to the storage means 40. And the repair processing result storage means 33 are stored in an appropriate database means 25.
[0021]
The information in the repair processing result storage means 33 is notified to the fail reading means 22, and thereafter, a mask is applied so that the fail element in the memory to be inspected having the same address as the repaired address is not read. It is also desirable that they are configured in such a manner.
[0022]
That is, in the repair operation processing apparatus for a semiconductor memory according to the present invention, in regard to the redundancy operation in the repair process for manufacturing the semiconductor memory, fail information is stored in a list-type data structure, and the pattern is processed into a pseudo segment. For the basic operation, a calculation processing flow as shown in FIG. 2 is adopted as an example.
[0023]
If the repair operation processing device 100 for a semiconductor memory according to the present invention is described in more detail, the fail reading means 22 executes a fail information receiving process S1 shown in the flowchart in FIG.
[0024]
That is, the fail reading means 22 scans the fail memory 21 storing the result of the fail inspection of the appropriate tester device, and fails information such as the address of the fail element determined to be failed, for example, as shown in FIG. Is sent to the list creation unit 23.
[0025]
Further, the fail reading means 22 has a function of receiving the repair confirmation information from the list creating means 23 and not continuously reading the address.
[0026]
In the list creation unit 23, the processing from the failure information creation processing S2 to the failure end determination processing S6 shown in the flowchart in FIG. 2 is performed.
[0027]
That is, in the fail information creating process S2, the fail information received from the fail reading means 22 is added to the fail information in the database 25.
[0028]
On the other hand, in the pseudo segment & list creation processing S3 shown in the flowchart in FIG. 2, each table of the pseudo segment 30, the fail list 32, and the fail information 31 is created.
[0029]
That is, the address on the fail list table 32 is compared with the address of the new fail element. If the address is a new address, the address is added to the corresponding list. If the address is existing, the list is newly created.
[0030]
FIG. 3 shows an example of the basic structure of a table included in the database 25 used in the present invention. The pseudo segment table 30 isAs shown in FIG.It represents a plurality of pseudo segments 51 to 54 composed of a fail pattern, and has a number for the fail list 32.
[0031]
Here, the semiconductor memory according to the present inventionIn the repair operation method, as described above, the first step of extracting the fail address of the fail element included in the semiconductor memory and one of the extracted fail addresses selected from the X address and the Y address are the same. Extract a certain fail element Two Step, a third step of extracting a fail element having the same address as the fail element extracted in the second step, and a fail element extracted in the third step, wherein the one address is the same as the fail element extracted in the third step. A fourth step of extracting a certain fail element, a fifth step of repeating the second to fourth steps until there is no more fail element to be extracted, and forming a pseudo segment group including the extracted fail element; A sixth step of allocating a spare line provided in the semiconductor memory in advance for each pseudo segment group, and a seventh step of determining whether there is a fail element to which the spare line is not allocated and whether there is an unused spare line And a repair operation method for a semiconductor memory including In the calculation method, the pseudo-segment groups are, for example, be formed by a method as shown below.
[0032]
That is, as shown in FIG. 4, a plurality of fail elements which are simultaneously generated on the same X-direction line or the same Y-direction line on the address are preferentially selected from all the fail elements. It is configured to be grouped.
[0033]
For example, it is detected whether or not a fail element exists for each of the Y direction lines in the X direction. If a fail element exists, a search is performed in the Y line direction where the fail element exists. Then, it is detected whether or not another fail element exists on the Y line.
[0034]
Then, based on the detection result, the relationship between the respective fail elements is determined.
[0035]
In other words, the pseudo segment group according to the present invention includes a plurality of pseudo segment groups which are highly correlated with each other from the addresses of all the fail elements in the failed state in the memory under test. To form.
[0036]
Specifically, when forming the fail information table according to the present invention, for example, first, the address is fixed in the Y direction, that is, fixed to (0), and the search is performed in the X address direction.
[0037]
Next, after searching for the Y line (0) to the end in the X direction, next, the next Y line, that is, the Y (1) line, is similarly searched in the X direction. is there.
[0038]
By repeating such an operation, a fail information table 10 as shown in FIG. 4 is completed.
[0039]
Next, when the pseudo segment table 12 as shown in FIG. 5 is formed from the fail information table 10, for example, in the fail information table 10, the address is first fixed in the Y direction, and Search in the direction of the address.
[0040]
Thereafter, it is also detected whether another fail element exists at the same time for the Y direction line where the fail element exists.
[0041]
Next, of all the fail elements in FIG. 4, for example, a shaded fail element group is one of the four fail elements existing in the X direction and two more in the Y direction. Since there are a number of fail elements, it is determined that these fail elements have some sort of strong relationship with each other. The pseudo segment groups 51 are formed by concentrating them at the places.
[0042]
Similarly, since the fail elements in Y (4) and Y (19) in the Y line can be determined to have a strong relationship with each other, they are concentrated in one place, and the pseudo segment group is concentrated. 52 is formed.
[0043]
Further, since the fail element at the address of X = CF and Y = ED is not recognized as having any strong relation with other fail elements, the fail element alone forms the pseudo segment group 53. .
[0044]
Similarly, since the fail elements on the Y (EF) and Y (FF) lines can be determined to have a strong relationship with each other, they are concentrated at one place to form the pseudo segment group 54.
[0045]
When the list is repeatedly created as described above, the pseudo segment table 12 shown in FIG. 5 is formed.
[0046]
In FIG. 5, the pseudo segment No. 1 corresponds to the pseudo segment 51, and the pseudo segment No. 2 corresponds to the pseudo segment 52. FIG. 6 shows the list structure in this state.
[0047]
That is, FIG. 6 constitutes the segment group shown in FIG.Faillist andThe relationship of each fail elementWith a pointerFIG.
[0048]
In the present invention, in the pseudo segment & list creation processing S3, various tables are generated from the pseudo segment table 12 shown in FIG. 5, that is, the pseudo segment table 14, the fail information table 15, the X and Y list tables. 16 and 17 are newly created.
[0049]
The pseudo segment list 30 according to the present invention includes, for example, the X list No., the Y list No., the number of X lists, and the number of Y lists as shown in the pseudo segment table 14 of FIG. Further, as shown in the fail information table 15 of FIG. 7B, the fail information table 32 includes, for example, the addresses of the individual fail elements in the failed state in the memory to be inspected, the pseudo segment No. It is preferable that the next list No. having the function as the pointer described above is included.
[0050]
Further, as shown in FIGS. 7C and 7D, the X and Y list tables 16 and 17 have addresses in the X and Y directions, a head fail number, a fail number, and a pointer, respectively. It is preferable that the next X list table or the next Y list table No. having the function described above is included.
[0051]
Here, first, an example of a method of creating the fail information table 15 used in the present invention will be described. First, a fail element (13, 0) at the upper left of the pseudo segment table 12 in FIG. 5 is selected. Is entered in the column of fail No. 1 of the fail information table 15, and since the fail element (13, 0) exists in the group of the pseudo segment No. 1, enter 1 in the column of the pseudo segment No. Then, (0, 2) is written in the next No. column indicating how the fail element (13, 0) has a positional relationship with another fail element.
[0052]
This is because, as shown in FIG. 5, the address of the other fail element connected to the Y line of the same address of the fail element (13, 0) is (6E, 0). Nothing is connected on the X line.
[0053]
Therefore, on the Y line of the same address, the failure No. 2In the columnThe numerical value 2 of the fail No. 2 of the fail element (6E, 0) to be written is written, while 0 is written on the X line of the same address because nothing is connected (0, 2). ).
[0054]
Similarly, the fail element (6E, 0) is entered in the address column of fail No. 2 in the fail information table 15, and the fail element (6E, 0) exists in the pseudo segment No. 1 group. , 1 is written in the column of the pseudo segment No., and the fail element (6E, 0) is connected to the fail element (6F, 0) on the Y line of the same address, and on the X line of the same address. Is not connected, the fail No. of the fail element (6F, 0) is adopted as 3, and (0, 3) is entered.
[0055]
On the other hand, when the next fail element (6F, 0) is selected, the fail element (6F, 0) is entered in the address column of fail No. 3 of the fail information table 15, and the fail element (6F, 0) is entered. , 0) exist in the group of the pseudo segment No. 1, so enter 1 in the column of the pseudo segment No., and furthermore, the fail element (6F, 0) sets the fail element (AB) on the Y line of the same address. , 0) and the fail element (6F, 80) on the X line of the same address. Therefore, the fail No. corresponds to fail No. 4 indicating the fail element (AB, 0). Is selected, and 9 which is the value of fail No 9 indicating the fail element (6F, 80) is selected and written (9, 4).
[0056]
The fail information table 15 is completed by the method as described above.
[0057]
If the configuration of the pseudo segment table 14 is described, the pseudo segment table No. 1 record is assigned to the group of the segment 51 in the pseudo segment 12 from the arrangement address of the fail element of the pseudo segment 12 in the memory under test. X list No. of each field of the pseudo segment table No. 1 record is 1 (corresponding to list No. 1 of X list table 16), Y list No. is 1 (corresponding to list No. 1 of Y list table 17), X list Let the number be 4 and the number of Y lists be 3.
[0058]
Note that the size of each pseudo segment is represented by the number of X lists and the number of Y lists in the pseudo segment table 14.
[0059]
The pseudo segment No. 2 in the pseudo segment table 14 indicates information on the fail element of the pseudo segment group 52 in relation to other tables, and the X list of each field of the pseudo segment table No. 2 record No. is 5, Y list No. is 2, X list number is 3, and Y list number is 2.
[0060]
This means that the first fail element (4F, 4) in the pseudo segment table No. 2 record corresponds to fail No. 5 in the fail information table 15, and the X address (4F) of the list No. 5 record in the X list table 16 And the Y address of the list No. 2 record in the Y list table 17(4)It corresponds to.
[0061]
Hereinafter, the pseudo segment table 14 is formed in a similar manner.
[0062]
Regarding the X and Y list tables 16 and 17, first, in the list No. 1 record of the X list table 16, the X address is (13), the first fail number is 1, the fail number is 1, and the next number is 0.
[0063]
On the other hand, in the list No. 1 record of the Y list table 17, the Y address is (0), the first fail number is 1, the fail number is 4, and the next fail is 4. This next No is a pointer indicating a list in the same pseudo segment.
[0064]
That is, the list No. 1 record in the X list table 16 corresponds to X list No. 1 in the pseudo segment table 14, and the failed element in the failed state in the pseudo segment group 1 is inspected. The information relating to the X address of the address (13, 0) in the memory is configured to be related to the information in another table, and the leading fail No. in the fail information table 15 is specified. The failure number indicates failure No. 1, and the number of failures indicates the number of fail elements viewed from the fail element (13, 0) on the X line at the same address in FIG. In the specific example of the present case, since there is only one fail element at the address of (13, 0), the column of the fail number is 1.
[0065]
In the next No, since the address of the next fail element connected on the Y line of the same address of the fail element is determined to be fail No 2 from the list of the fail information table 15, Fail No. 2 is described.
[0066]
Hereinafter, the X list table 16 is configured based on the same concept.
[0067]
On the other hand, FIG.(D)List No. 1 record of the Y list table 17 corresponds to Y list No. 1 in the pseudo segment table 14, and the address (13) of the fail element in the pseudo segment group 1 in the memory under test. , 0), the information relating to the Y address is configured to be associated with the information in another table, and the leading fail No. is fail No. 1 in the fail information table 15. The number of failures indicates the number of fail elements viewed on the Y line of the same address from the fail element (13, 0) in FIG. 5, and in the present specific example, Since four fail elements are continuously present in the right direction from the 0 address of (13, 0) in the fail element, the column of the number of failures is four. You have me.
[0068]
In the next No, since it is known from the list of the fail information table 15 that the address of the next fail element connected on the Y line of the same address of the fail element corresponds to the fail No. 4, Fail No. 4 is described.
[0069]
Hereinafter, the Y list table 17 is configured based on the same concept.
[0070]
On the other hand,Fail information table 15In the column of pseudo segment No., 1 is entered to indicate that the fail element (13, 0) of fail No. 1 exists in pseudo segment No. 1, that is, in the pseudo segment group 51.
[0071]
Similarly, each of the fail element (6E, 0) of the fail No. 2, the fail element (6F, 0) of the fail No. 3 and the fail element (AB, 0) of the fail No. 4 are included in the pseudo segment No. 1, that is, in the pseudo segment group 51. Each one is entered to indicate that it exists.
[0072]
In addition, 2 is entered to indicate that the fail element (4F, 4) of fail No. 5 exists in the pseudo segment No. 2, that is, in the pseudo segment group 52.
[0073]
FIG. 3 shows an example of the basic structure of a table included in the database 25 used in the present invention. The pseudo segment table 30 represents a plurality of pseudo segments 51 to 54 formed by a fail pattern, and has a number for the fail list 32.
[0074]
The example of FIG. 7 is obtained by adding auxiliary data fields such as the number of lists to the basic configuration shown in FIG. 3 as described above.
[0075]
Here, the procedure for forming the various tables described above will be described more specifically.
[0076]
That is, each of the above tables may be a table in which the fail information table 15 is created in advance and another table is created based on the table. May be sequentially formed.
[0077]
That is, the address (13, 0) of the first fail element in the memory to be inspected is used for the fail information receiving process.S1As shown in FIG. 8B, in the fail information creation process S2, the address (13, 0) is entered in the address column and the address (13, 0) is entered in the next column of the fail information table, as shown in FIG. Add.
[0078]
Fail No is defined as starting from 1, and the next No (0, 0) is the next for both X and YfailIndicates that there is no pointer leading to.
[0079]
At this stage, the address of the adjacent fail element connected to the fail element (13, 0) next on the Y line of the same address and on the X line of the same address is not determined. (0, 0) is entered, and the pseudo segment group No. of the fail element (13, 0) is not yet determined, so 0 is entered.
[0080]
On the other hand, in the X list table 16 shown in FIG. 8C, referring to the fail information table of FIG. Corresponding 1 is entered, 13 is entered in the X address column, 1 is entered in the column of the first fail No., 1 is entered in the column of the number of fail, and any fail element is entered in the next No. column. Since it is connected or undecided, 0 is entered.
Also, as shown in FIG.In the Y list tableReferring to the fail information table of FIG. 8B, 1 corresponding to the first column of the fail information table 15 is entered in the list No. column, and 0 is set in the Y address column. In the column of fail No., 1 is entered in the column of fail number, and in the next No. column, 0 is entered because it is undetermined which fail element is to be connected.
[0081]
Then, in the pseudo segment table 14, 1 corresponding to the first column of the fail information table of FIG. 8B is written in the X list No. and Y list No. columns, and the X list number column And 1 in the Y list number column.
[0082]
After that, when the second fail element (6E, 0) is detected, as shown in FIG.Fail No2The address (6E, 0) is added to the address column of the record, and (0, 0) is added to the next No column.
[0083]
At this point, the fail element (13, 0) is connected to the fail element of the fail element (6E, 0) on the Y line of the same address, and on the X line of the same address. Since nothing is connected, the column for the next No in the fail No 1 record is rewritten as (0, 2).
[0084]
In this case, 2 corresponds to 2 of fail No. 2 in the fail information table 15.
[0085]
Since the fail No. 2 is added to the fail information table 15, the X list table 16 shown in FIG. 9C refers to the fail information table of FIG. In the column, 2 corresponding to the second column of the fail information table 15 is entered, 6E is entered in the X address column, 2 is entered in the first fail No column, and 1 is entered in the fail count column. In the next No column, it is undecided what fail element is to be connected, so 0 will be entered. However, the value of 0 in the next No column in the above list No. 1 Will be rewritten to 2.
[0086]
Further, in the Y list table 17 shown in FIG. 9D, referring to the fail information table of FIG. 9B, since the address on the X line of the same address has not changed, the list No. The column remains at 1, the Y address column remains at 0, and the head fail No column remains at 1.
[0087]
On the other hand, in the field of the number of failures, 2 is entered because the second failure element is arranged on the Y line of the same address.
[0088]
Then, in the pseudo segment table 14 shown in FIG. 9A, the value of 1 in the X list No. and Y list No. columns does not change, the X list number column is rewritten from 1 to 2, and the Y list number column A value of 1 means no change.
[0089]
By repeatedly executing the above operation, the respective tables 14 to 17 shown in FIGS. 7A to 7D are completed.
[0090]
In this case, the X list No. of each field of the No. 1 record in the pseudo segment table 14 is 1, the Y list No. is 1, the number of X lists is 4, and the number of Y lists is 3.
[0091]
Next, when the fail element (6E, 0) is received, the address (6E, 0) and the next No (0, 3) are added to the No. 2 record of the fail information table 15 as in the first fail processing. The procedure in that case will be described in more detail below.
[0092]
That is, since there is no list having the address (6E) in the X list table, the address (6E) and the like are added to the list No. 2 record.
[0093]
Next, a search of the Y list table finds the list No. 1 having the same Y address (0) as the fail (6E, 0), so that list addition processing and pseudo segment processing are performed in the following procedure.
[0094]
(Step 1): In the found table, the fail number field of the corresponding record (list No. 1) is incremented to 2.
[0095]
(Step 2): The first fail No. 1 of the list No. 1 record is searched from the fail information table, and its address (13, 0) is obtained.
[0096]
In order to create a list in the Y direction, fail No. 2 of fail address (13, 0) is added to the next No. to be (0, 2).
[0097]
(Step 3): From the pseudo segment No. 1 of the fail No. 1, the number of X lists of the No. 1 record in the pseudo segment table is incremented to 2.
[0098]
(Step 4): From the No. 1 record of the pseudo segment table, it can be seen that the X list No. starts from 1, so the next No. of the list No. 1 record on the X list table is 2.
[0099]
Similarly, when the next fail element (6F, 0) is received, the address (6E, 0) and the next No (9, 4) are added to the No. 3 record of the fail information table as in the first fail processing. .
[0100]
By repeating the above operation, a series of tables is completed except for the pseudo segment No. column in the fail information table.
[0101]
As described above, the X list table 32-1 and the Y list table 32-2 that configure the fail list 32 have the same address.ComposedHas list information.
[0102]
On the other hand, the fail information table 31 has pointer information (for example, the next No (X, Y) as shown in FIG. 7) indicating a fail address or a connection to a fail element having the same address.
[0103]
In the address determination processing S4 shown in the flowchart of FIG. 2, the determination is made based on the number of spare lines and the like as a criterion.
[0104]
Further, in the address determination determination processing S4 shown in the flowchart in FIG. 2, the notification processing S5 to the pair reading means (not shown) is performed. To the confirmed spare line and save the result.
[0105]
If the address is not determined in the address determination determination processing S4, the read notification processing S5 is passed, and the process proceeds to step S6 for determining whether or not the failure is completed.
[0106]
In the fail end determination processing S6, communication is performed with the reading means 22 for unread fail information. The process of each step is repeatedly executed, and thereafter, it is determined again whether there is no unread fail information. If all the fail information is read, the process proceeds to a redundancy operation described later.
[0107]
The redundancy calculation means 24 executes the processing from the fixed line search processing S7 to the spare remaining determination processing 12.
[0108]
In the fixed line search process S7, the spare line to be fixed and its address are determined with reference to the pseudo segment table 30 and the fail list table 31.
Further, it has a function of comparing the number of pseudo segments in the segment with the number of remaining spare lines to determine whether repair is impossible. In the fail information updating process S8, a decision flag is set for the decided line or fail element.
[0109]
In the present invention, a specific method of comparing the number of pseudo segments with respect to the fail element and the number of remaining spare lines to determine whether or not repair is possible can use a conventionally known technique.
[0110]
In the pseudo segment & list update processing S9, the list information of each table is updated. In the confirmed line storage processing S10, the confirmed spare line and its address are notified to the database 25.
[0111]
In the unprocessed failure determination process S11, the process ends if all the fail elements have been rescued. In the spare remaining determination process S12, the process ends when there is no spare line.
[0112]
As described above, when the semiconductor memory repair operation processing device 100 according to the present invention is used to execute the semiconductor memory repair operation method, as shown in FIG. A fail information table 10 for storing information such as an address of a fail element in a memory to be inspected is created. AndPseudo segmentIn the & list creation process S3, the pseudo segment table 12 expressing the arrangement of the fail elements and the X and Y list tables 16 and 17 as shown in FIGS. 7C and 7D are compared with the existing fail list. create. Further, in the redundancy operation processing from the fixed line search processing S7 to the fixed line storage processing S10, an address for using the spare line is determined with reference to each table, and a repair solution for the fail element is obtained. As a result, it is possible to reduce the storage area of the fail information, speed up the arithmetic processing, and improve the rescue rate.
[0113]
Here, when individually comparing each pseudo segment group of the pseudo segment table groups 51 to 54 used in the present invention with the number of spare lines provided in advance in the X and Y directions, It is preferable that the control is performed so that the pseudo-segment groups are sequentially selected in descending order of the size and the comparison processing with the number of spare lines is executed.
[0114]
In the present invention, the pseudo segment table 14, the fail information table 15, the X and Y list tables 16 and 17 are subjected to the comparison processing, and the repair processing of a predetermined fail element group is determined. It is also desirable to provide an updating unit 27 for performing an updating process on each of the remaining fail elements so that the latest information is displayed.
[0115]
Next, the processing of the redundancy operation unit will be described using an example of a database (FIG. 7) and an example of a repair result (FIG. 10).
[0116]
In the fixed line search process 7, the list having the largest number of fail elements in the pseudo segment having the largest size (sum of the number of lists in the X direction and the Y direction) is selected. According to the search of the pseudo segment table, the segment No. 1 is the largest in size 7.
[0117]
Of the lists among them, the list having the largest number of fail elements is selected as the Y address (0) of list No1 having the number of fails of 4 by searching the Y list table (FIG. 10, fixed (1) ).
[0118]
In the fail information updating process 8, processes such as setting a decision flag on a decided line or a rescued fail element are performed.
[0119]
Regarding the fail information table,FIG. 11 (B)As shown in the figure, the pseudo segment No. including the fail Nos. 1 to 4 is set to 0, and the next No. indicating the list is also set to (0, 0), indicating that the rescue is performed.
[0120]
In the pseudo segment & list update processing S8, the table relating to the determined spare line is updated. Y address (0)Spare lineWas confirmed,FIG. 11 (A)As shown in the pseudo segment table,X list No. of the No. 1 segment is 3, Y list No. is 4., The number of X lists is 1, and the number of Y lists is 2. Further, with respect to the Y list table, the determined list No. 1 record is deleted, and further, with respect to the X list table, the list related to the rescued fail element is updated.
[0121]
In the confirmed line storage processing S10, processing related to the spare line, such as storage of the spare line information whose use is confirmed and decrement of the available spare line, is performed.
[0122]
Since there is an unprocessed fail element in the unprocessed failure judgment processing S11 and there is a spare line remaining in the spare remaining judgment processing S12, the following list processing is performed.
[0123]
In the next confirmed line search process S7, the spare line of the Y address (4) is selected because the size of the segment 52 is the maximum at the size 6 (determined (2)).
[0124]
However, in this example, when the maximum number of X and Y lists is the same, Y is set.
[0125]
When the arithmetic processing is sequentially performed, the result up to the determination (7) in FIG. 10 is obtained.
[0126]
Finally, FIG. 12 shows an example of the database immediately before the failure No. 12 is determined.
[0127]
In the confirmed line search process 7 of the operation process flow of FIG. 2, the No. 4 segment of the pseudo segment table is selected, and the fail information updating process S8 and thereafter are executed.
[0128]
As a result,Segment tableAre deleted and the other tables are updated.
[0129]
In the unprocessed fail determination process S11, since there is no data in the pseudo segment table, it is determined that all fail elements have been rescued, and the calculation process ends.
[0130]
More specifically, according to the present invention, as an example of a plurality of segments shown in FIG. 7, a fixed line search process 7 is performed by adding the following process. The sum of all pseudo-segments in the segment and the remainingSpare lineIs equal to the total number ofallThe fail element is relieved first.
[0131]
Also, the sum of the short list of pseudo-segments in a segment is the minimum (requirement) for that segment's required spare line, so the remainingSpare lineIf the number is larger than the above, it can be determined that relief is impossible.
[0132]
This processing makes it possible to succeed in the rescue as shown in FIG. 10, and to speed up the determination of the rescue impossible.
[0133]
As described above, the semiconductor memory repair operation method according to the present invention basically employs the following technical configuration. That is, a repair operation method in a repair processing step in a semiconductor memory manufacturing process, wherein all fail elements in a failed state are selectively extracted from the detected fail information of the memory, and the selected and extracted From the addresses of all the fail elements in the memory under test, a plurality of pseudo segment groups each including a plurality of the fail elements having high relevance are formed, and a pseudo segment table is created. Ensuring relevance on individual addresses of fail elementsIn order toA fail information table and an X and Y list table are created, and each pseudo segment group of the pseudo segment table group is individually compared with the number of spare lines provided in advance in the X and Y directions. This is a repair operation method for a semiconductor memory configured to determine whether it is possible or not.
[0134]
In the repair operation method of the semiconductor memory according to the present invention, the pseudo segment group may include the same X-direction line on the address of the memory under test from all the fail elements. Alternatively, it is desirable that a plurality of fail elements occurring simultaneously be preferentially grouped in the same line in the Y direction.
[0135]
Similarly, in the repair operation method of the semiconductor memory according to the present invention, each pseudo segment group of the pseudo segment table group is individually compared with the number of spare lines provided in advance in the X and Y directions. In doing so, it is also preferable that the individual pseudo segment groups are selected in descending order of size and the comparison process with the number of spare lines is executed.
[0136]
On the other hand, in the repair operation method of the semiconductor memory according to the present invention, the pseudo segment table, the fail information table, and the X and Y list tables are stored in the predetermined fail element group after the comparison process is executed. After the repair process is determined, it is also preferable that the update process is executed for each of the remaining fail elements so that the latest information is displayed.
[0137]
【The invention's effect】
As described above, the semiconductor memory repair operation method and the semiconductor memory repair operation processing device of the present invention employ the above-described technical configuration, so that the speed of the arithmetic processing is reduced and the memory for storing the fail information is reduced. It becomes possible. For example, for speeding up, when the Y address (4) is determined, the process of extracting the relevant fail element is performed using the conventional method, and the search from the address (13) to (7F) is performed seven times. However, the present invention can be implemented only by searching the list from the Y address (4) twice.
[0138]
In addition, as for the reduction of the storage memory, as shown in FIG. 4, a 10 × 8 table is required even when 14 fail storages are used, as shown in FIG. ing. According to the present invention, the storage area is sufficient for only the fail existing in the list type structure shown in FIG. 6 in the database shown in FIG.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a specific configuration of a repair operation processing device for a semiconductor memory according to the present invention.
FIG. 2 is a flowchart showing a specific operation procedure of a repair operation method for a semiconductor memory according to the present invention;
FIG. 3 is a diagram showing an example of a basic configuration of various data tables used in the present invention.
FIG. 4 is a diagram showing an example of fail information used in the present invention.
FIG. 5 is a diagram showing an example of a pseudo segment table used in the present invention.
FIG. 6 is a list showing an association of each fail element in a pseudo segment table used in the present invention.
FIG. 7 is a diagram showing an example of a data configuration of various data tables used in the present invention.
FIG. 8 is a diagram showing a change state of various data tables according to the present invention.
FIG. 9 is a diagram showing a change state of various data tables in the present invention.
FIG. 10 is a diagram illustrating an example of a result after a repair process according to the present invention.
FIG. 11 is a diagram showing a state of change of various data tables in the present invention.
FIG. 12 is a diagram showing a change state of various data tables according to the present invention.
FIG. 13 is a diagram illustrating an example of successful repair processing when a plurality of segments exist.
FIG. 14 is a diagram illustrating an example in which repair processing has failed when a plurality of segments exist.
FIG. 15 is a flowchart illustrating an example of an operation method of a conventional repair operation method of a semiconductor memory.
FIG. 16 is a diagram showing an example of a data structure used in a conventional semiconductor memory repair operation method.
[Explanation of symbols]
10 ... Fail information table
12 ... Pseudo segment table
14 ... Pseudo segment table
15 ... Fail information table
16, 17 ... X, Y list table
21 ... Fail memory
22 ... Fail reading means
23 ... List creation means
24 ... Repair processing means
25 ... Database means
27 ... Update means
30 ... Pseudo segment table
31 ... Fail information table
32 ... X, Y list table
32-1 ... X-direction list table
32-2... Y-direction list table
33 ... Repair processing result storage means
40 ... storage means
51 to 54: pseudo segment
100: repair operation processing device for semiconductor memory

Claims (13)

A first step of extracting a fail address of a fail element included in the semiconductor memory;
A second step of extracting, from among the extracted fail addresses , a fail element having one of the same addresses selected from an X address and a Y address ;
A third step of extracting a fail element having the same address as the fail element extracted in the second step ;
A fourth step of extracting a fail element having the same address as the fail element extracted in the third step ;
The second to fourth steps are repeated until there are no more fail elements to be extracted, and the fifth step is to form a pseudo segment group including the extracted fail elements.
Steps and
A sixth step of allocating a spare line provided in advance in the semiconductor memory for each of the pseudo segment groups;
Repair operation method of a semiconductor memory which comprises a seventh step of determining the presence or absence of the spare line presence and unused fail element the spare line is not assigned. 2. The repair operation method for a semiconductor memory according to claim 1, wherein in the sixth step , the spare lines are allocated in an order of increasing number of fail elements included. 3. The repair operation of the semiconductor memory according to claim 2, wherein in the sixth step , an X-direction line or a Y-direction line having the largest number of fail elements is selected first and the spare line is allocated. Method. In the sixth step , when judging the number of fail elements included in the pseudo segment group or the number of fail elements included in the X-direction line or the Y-direction line, a fail element to which a spare line has already been assigned 4. The repair operation method for a semiconductor memory according to claim 3, wherein the number is not included. The method according to claim 1, wherein the sixth step and the seventh step are repeated until one of the fail element to which the spare line is not assigned and the unused spare line disappears. 5. A repair operation method for a semiconductor memory according to any one of the above. Extracting a fail address of a fail element included in a semiconductor memory, extracting the fail address, and a pseudo-segment number commonly assigned to a fail element in which one of an X address and a Y address is the same among the fail addresses; A fail information table containing pointer information indicating a connection between fail elements to which the same pseudo segment number is allocated;
An X address of a first detected X fail element detected first among fail elements having the same X address;
A first fail number indicating where the information of the first detected X fail element is described in the fail information table;
Fail count indicating the number of fail elements having the same X address as the first detected X fail element;
First among the fail elements belonging to the same pseudo segment group as the first detection X fail element and belonging to the X address of the first detection X fail element and the adjacent X address in the pseudo segment group, the first step An X list table including a fail element detected in the first list and a next list number indicating where the fail element detected after the first detection X fail element is described in the table.
A Y address of a first detected Y fail element detected first among fail elements having the same Y address;
A leading fail number indicating where the information of the first detected Y fail element is described in the fail information table;
Fail count indicating the number of fail elements having the same Y address as the first detected Y fail element;
First among the fail elements belonging to the same pseudo segment group as the first detected Y fail element and belonging to the Y address of the first detected Y fail element and the adjacent Y address in the pseudo segment group, the first step A fail element detected in the Y-list table including a next list number indicating where the fail element detected after the first detected Y fail element is described in the same table,
Where in the X list table and the Y list table, the information of the fail element detected first among the failures in the pseudo segment group composed of the fail elements to which the common pseudo segment number is allocated is described. A pseudo segment table including an X list number and a Y list number indicating
A first step of creating
Using the X list number, the Y list number, and the next list number, the X direction line and the Y direction line included in the same pseudo segment group are described in the X list table and the Y list table. The number of fail elements included in each of the pseudo segment groups is counted by referring to the number of failed elements, and the number of fail elements included in each of the pseudo segment groups is larger in the pseudo segment group order, and is provided in advance in the semiconductor memory. A second step of allocating a spare line,
Determining whether or not there is a fail element to which the spare line is not allocated and whether or not there is an unused spare line. 2. The method according to claim 1, wherein, in the second step, the spare line is allocated by first selecting an X-direction line or a Y-direction line having the largest number of fail elements by using the referenced fail number. 7. The repair operation method for a semiconductor memory according to item 6. After the second step is executed and the spare line is allocated, the latest information on the remaining fail elements is displayed in each of the fail information table, the X list table, the Y list table, and the pseudo segment table. The method according to claim 7, further comprising a fourth step of executing a process. 9. The method according to claim 6, wherein the second step and the third step are repeated until one of the fail element to which the spare line is not assigned and the unused spare line is eliminated. A repair operation method for a semiconductor memory according to claim 1. A repair operation processing device for a semiconductor memory, wherein fail read means for outputting fail information obtained by detecting the presence or absence of a failure for each memory from the semiconductor memory;
A fail element in which a fail has occurred is extracted from the fail information, and a pseudo fail group consisting of fail elements having one of the same X address or Y address among the addresses in the semiconductor memory is formed. List creating means for creating a fail information table including an address of the element in the semiconductor element and information indicating which of the pseudo fail groups the fail element belongs to,
Storage means for storing the fail information table;
Repair processing means for allocating a spare line provided in advance in the semiconductor memory for each of the pseudo segment groups in accordance with the fail information table stored in the storage means. Processing equipment. The repair processing means selects the segment group in which the number of fail elements included is large in order, selects a line in which the number of fail elements included is the largest in the selected pseudo segment group, and executes the processing. The repair operation processing device for a semiconductor memory according to claim 10, wherein: After the repair processing means completes the assignment of a spare line to a predetermined fail element, updating means for updating the information displayed in the fail information table to the latest information on the remaining fail elements is further provided. The repair operation device for a semiconductor memory according to claim 10, wherein: 13. The fail information table according to claim 10, wherein the fail information table further includes pointer information indicating a connection to a fail element having one of the same X address or Y address as the fail element. The repair operation device for a semiconductor memory according to any one of the above.

Images