CN-121984514-A - Semiconductor data processing method, device, computer equipment and storage medium

CN121984514ACN 121984514 ACN121984514 ACN 121984514ACN-121984514-A

Abstract

The invention relates to a semiconductor data processing method, a device, computer equipment and a storage medium, which comprise the steps of traversing WIP source data, carrying out differential coding on different batch IDs and corresponding wafer IDs according to coding types to generate batch codes and wafer codes, compressing at least one wafer code in the same data dimension to generate target compression data comprising batch codes, the wafer codes and/or continuous compression values, wherein the batch codes are obtained according to the wafer codes of the full batches, and the continuous compression values are obtained according to continuous batch codes and compression rules or conversion of continuous wafer codes and compression rules. By adopting the data processing method, the memory storage of the data can be reduced and the data query efficiency can be provided.

Inventors

Jin Peiyuan
CHEN XIAOCHUAN
SHAO KANGPENG

Assignees

杭州广立微电子股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260126
Priority Date: 20251231

Claims (14)

1. A method of processing semiconductor data, comprising: traversing WIP source data, performing differential coding on different lot IDs and corresponding wafer IDs according to coding types, and generating a lot code and a wafer code, wherein the same wafer ID of the same lot corresponds to the same wafer code; and compressing at least one wafer code under the same data dimension to generate target compressed data comprising a batch code, the wafer code and/or a continuous compression value, wherein the batch code is obtained by compressing the wafer code of a full batch, and the continuous compression value is obtained by converting the continuous batch code and the compression rule or the continuous wafer code and the compression rule.
2. The method of claim 1, wherein said compressing at least one of said wafer encodings in the same data dimension generates target compressed data comprising batch encodings, wafer encodings, and/or consecutive compressed values, comprising: performing full batch compression on at least one wafer code under the same data dimension to generate first compressed data comprising batch codes and/or wafer codes; and continuously encoding and compressing the first compressed data to generate target compressed data comprising batch codes, wafer codes and/or continuous compression values.
3. The method of claim 2, wherein compressing at least one of the wafer encodings in the same data dimension generates first compressed data including a lot encoding and/or a wafer encoding, comprising: judging whether the lot codes under the same data dimension exist the wafer codes of the full lot or not; If the wafer codes of the full batch exist, recording the wafer codes of the full batch by using the wafer codes of the full batch, and recording the rest wafer codes; and generating first compressed data according to the recorded batch code and/or the wafer code.
4. A semiconductor data processing method according to claim 3, wherein the pre-statistics obtain a table of a number relationship between wafer IDs and wafer numbers thereof under the same lot ID under the same data dimension, and the compressing at least one wafer code under the same data dimension generates first compressed data including the lot code and/or the wafer code, and the method comprises: Presetting a wafer name list comprising full lot wafer IDs, a wafer set and the number of full lots; Sequentially traversing the wafer IDs of the wafer name list as current wafer IDs, comparing the current wafer IDs with the wafer IDs of the number relation table, recording the current wafer IDs through a wafer set if the same wafer IDs exist in the number relation table, and acquiring the number of wafers corresponding to the current wafer IDs to update the number of full batches, wherein the number of full batches only records the minimum number of wafers; When traversing is finished, judging whether the wafer set meets the size of a full batch, if so, recording batch codes corresponding to the batch IDs in the number of times equal to the number of the full batch, wherein each wafer ID update in the number relation table reduces the number of wafers equal to the number of the full batch; and obtaining the remaining wafer IDs and the number of the wafers in the number relation table, and recording the wafer codes corresponding to the wafer IDs in the number equal to the number of the remaining wafers.
5. The method of claim 1, wherein the differential code uses a positive/negative code type to generate the lot code and the wafer code.
6. The method of claim 1, wherein the continuous compression value is obtained according to a continuous batch coding and compression rule or a continuous wafer coding and compression rule conversion, comprising: judging whether continuous codes exist in the batch codes or the wafer codes; If continuous codes exist, generating target code values for the continuous batch codes and/or the wafer codes by adopting the target code rules; and converting according to the target coding value to obtain a continuous compression value.
7. The method for processing semiconductor data as recited in claim 6, wherein, the method for generating the target coding value by adopting the target coding rule comprises the following steps of: Respectively encoding numerical bits and continuous bits of the generated target encoding values according to the initial encoding and the continuous number of the continuous batch encoding and/or the wafer encoding; And encoding the sign bit according to whether the compressed code is batch code or wafer code or not and whether continuous compression is carried out, so as to generate a target code value.
8. The method according to claim 7, further comprising decompression processing of the target compressed data, comprising: obtaining a target value in target compressed data, judging whether the target value is a wafer code according to the code type of the target value, recording a wafer ID corresponding to the wafer code if the target value is the wafer code, and judging whether the target value is a continuous compressed value if the target value is an amorphous wafer code: when the target value is a continuous compression value, decoding to obtain a continuous batch code or wafer code, and obtaining and recording a corresponding batch ID or wafer ID according to the continuous batch code or wafer code; When the target value is discontinuously compressed, the representative target value is a lot code, and the wafer ID corresponding to the full lot is obtained and recorded according to the lot code.
9. The method according to claim 8, wherein the determining whether the target value is a continuously compressed value comprises: and converting the target value into a target coding value, acquiring a sign bit representing whether the target coding value is compressed, judging whether the sign bit is compressed data, and if the sign bit is the compressed data, obtaining the representing target value as a continuous compression value.
10. The method of claim 9, wherein decoding to obtain a continuous lot code or wafer code when the target value is a continuous compressed value, comprises: acquiring a sign bit representing the type of the compressed code in the target code value, and judging the target value as a batch code or a wafer code; And respectively acquiring the initial code and the continuous number according to the numerical bit and the continuous bit in the target code value, and selecting a circulation rule corresponding to the batch code or the wafer code to acquire the continuous batch code or the wafer code.
11. A semiconductor data processing apparatus, comprising: the code generation module is used for carrying out differential codes on different batch IDs and corresponding wafer IDs according to the code types to generate batch codes and wafer codes, wherein the same wafer ID of the same batch corresponds to the same wafer code; The code compression module is used for compressing at least one wafer code under the same data dimension to generate target compression data comprising a batch code, the wafer code and/or a continuous compression value, wherein the batch code is obtained according to the wafer code of a full batch, and the continuous compression value is obtained according to continuous batch code and compression rules or continuous wafer code and compression rules in a conversion mode.
12. The semiconductor data processing apparatus of claim 11, wherein the encoding compression module comprises: the full-batch code compression unit is used for carrying out full-batch compression on at least one wafer code under the same data dimension to generate first compression data comprising batch codes and/or wafer codes; And the encoding compression unit is connected for carrying out continuous encoding compression on the first compression data to generate target compression data comprising batch encoding, wafer encoding and/or continuous compression values.
13. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the semiconductor data processing method of any one of claims 1 to 10 when the computer program is executed.
14. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program implementing the steps of the semiconductor data processing method of any one of claims 1 to 10 when executed by a processor.

Description

Semiconductor data processing method, device, computer equipment and storage medium Technical Field The present invention relates to the field of semiconductor data processing, and in particular, to a semiconductor data processing method, apparatus, computer device, and storage medium. Background In the field of Integrated Circuit (IC) fabrication, WIP (Wafer In Process) is a core management object throughout the whole production Process, meaning all wafers In "unfinished state" from Wafer input fabrication (after dicing, grinding) to completion of all Front End (FEOL), back End (BEOL) processes and through Final Test (FT), waiting for shipment. The influence of factors such as a machine on wafer manufacturing can be reflected to a certain extent through the relevance analysis of WIP data and other data of different wafer testing steps, and a common mode adopted for acquiring the relevance data between the WIP data and the data of other different wafer testing steps is join logic at a database level, so that the efficiency of the mode is ideal when the WIP data volume is not large, but the data lines required to be scanned by the join logic are gradually increased along with the gradual increase of the data volume, and finally the SQL time consumption of the whole query is overlong and even is blocked. Disclosure of Invention The invention provides a semiconductor data processing method, a device, a computer device and a storage medium for solving all or part of the problems in the prior art, and can deal with the association inquiry of huge WIP data by carrying out special compression processing and decompression processing on the WIP data, thereby greatly improving the inquiry efficiency. In a first aspect, in this embodiment, there is provided a semiconductor data processing method including: traversing WIP source data, performing differential coding on different lot IDs and corresponding wafer IDs according to coding types, and generating a lot code and a wafer code, wherein the same wafer ID of the same lot corresponds to the same wafer code; and compressing at least one wafer code under the same data dimension to generate target compressed data comprising a batch code, the wafer code and/or a continuous compression value, wherein the batch code is obtained by compressing the wafer code of a full batch, and the continuous compression value is obtained by converting the continuous batch code and the compression rule or the continuous wafer code and the compression rule. In some of these embodiments, the compressing at least one of the wafer encodings in the same data dimension to generate target compressed data including a lot encoding, the wafer encoding, and/or a consecutive compressed value includes: performing full batch compression on at least one wafer code under the same data dimension to generate first compressed data comprising batch codes and/or wafer codes; and continuously encoding and compressing the first compressed data to generate target compressed data comprising batch codes, wafer codes and/or continuous compression values. In some of these embodiments, the compressing at least one of the wafer encodings in the same data dimension generates first compressed data including a lot encoding and/or a wafer encoding, including: judging whether the lot codes under the same data dimension exist the wafer codes of the full lot or not; If the wafer codes of the full batch exist, recording the wafer codes of the full batch by using the wafer codes of the full batch, and recording the rest wafer codes; and generating first compressed data according to the recorded batch code and/or the wafer code. In some embodiments, the pre-counting obtains a table of a number relationship between the wafer IDs and the number of wafers in the same lot ID in the same data dimension, and compresses at least one wafer code in the same data dimension to generate first compressed data including the lot code and/or the wafer code, where the first compressed data includes: Presetting a wafer name list comprising full lot wafer IDs, a wafer set and the number of full lots; Sequentially traversing the wafer IDs of the wafer name list as current wafer IDs, comparing the current wafer IDs with the wafer IDs of the number relation table, recording the current wafer IDs through a wafer set if the same wafer IDs exist in the number relation table, and acquiring the number of wafers corresponding to the current wafer IDs to update the number of full batches, wherein the number of full batches only records the minimum number of wafers; When traversing is finished, judging whether the wafer set meets the size of a full batch, if so, recording batch codes corresponding to the batch IDs in the number of times equal to the number of the full batch, wherein each wafer ID in the number relation table is updated to reduce the number of wafers equal to the number of the full batch; and obtaining the remaining wafer IDs and the number o