46 495 0 A7 _____B7 _ 五、發明說明（/ ) 技術領域： 本發明係關於一種微影機台載荷分配（all〇cati〇n)系統 之最佳化方法，特別是關於在各批次晶片進入各微影機台製 程之前’加以整合處理’使之分配最適合的微影機台載荷量， 以使適當之機台分配於各批次欲處理之晶片，亦可增加工作 效率並且使各機台之處理達到最佳平衡。 發明背景： 目刚積體電路製造技術（IC Manufacture technology)又 進入更複雜且要求精確之製程，而其中微影黃光 (Photolithography)製程又位居整個半導體製程之關鍵地 位，除了微影技術本身之外，微影機台本身亦是影響製程之 因素。在最小線寬（Feature Size) 0.30以m以上之積體電路 製轾，微影機台於定義不同層次時產生的微小差異現象，皆 疋在所此掌控之範圍内。然而，當進入〇Μ从m以下之製程， 疊對（Overlay)係為整個微影製程非常重要之步驟，因此大 部份的微影步驟，即使於相同機台製程，如步進機（stepper) 或掃瞄機（Scanner)，但由於各機台之製程特性之微小差異， 仍必須侷限在特定之機台執行，且其接續的每一站處理 (Stage)製程都必須與前述機台為搭配之機台 Tool) ’甚至須指定在同一機台加工，以使整個製程之動作精 確’並可付彳政影過程之最佳疊對控制（QVerlay Consol )結果。 因此，在第-層次（Layer)機台之選擇調度顯得特別重 要，其將景彡響後續整個微影製程的品質及時間控制，對製程 ___ 2 本紙張尺度適用中國β標準（CNS)A4規格（210 x 297公愛^--—-- (請先閱讀背面之注意事項再填寫本頁)46 495 0 A7 _____B7 _ V. Description of the invention (/) Technical field: The present invention relates to an optimization method for a lithography machine load distribution (all〇cati〇n) system, especially regarding the entry of wafers in each batch. 'Integrate processing' before each lithography machine process to distribute the most appropriate lithography machine load, so that the appropriate machine is allocated to the wafers to be processed in each batch, which can also increase work efficiency and make each machine The processing of the stage reaches the best balance. Background of the Invention: IC Manufacture technology has entered a more complex and precise manufacturing process, and the photolithography process is in a key position in the entire semiconductor manufacturing process, except for the lithography technology itself. In addition, the lithography machine itself is also a factor affecting the manufacturing process. When the minimum line width (feature size) is 0.30 and the integrated circuit is more than m, the micro-photographing machine will define the small differences in the different levels of the phenomenon, which are all within the control range. However, when entering a process of 0 μm and below, Overlay is a very important step for the entire lithography process. Therefore, most of the lithography steps are performed on the same machine, such as a stepper. ) Or Scanner, but due to the small differences in the process characteristics of each machine, it must still be limited to a specific machine to execute, and each subsequent stage processing (Stage) process must be the same as the aforementioned machine Tool Tool) 'Even it must be specified to be processed on the same machine to make the entire process accurate' and it can support the best QVerlay Consol result of the government shadow process. Therefore, the selection and scheduling of the first-level (Layer) machine is particularly important. It will affect the quality and time control of the subsequent lithography process, and the process ___ 2 This paper standard applies Chinese β standard (CNS) A4 Specifications (210 x 297 public love ^ ------) (Please read the precautions on the back before filling in this page)
464950 A7 ----- B7 五、發明說明（2) 懸度要求極高之半導體業而言，其乃不可忽視之問題。而 且，若於微影製程中選擇變換已指定之機台，將導致暴㈣ 制超it規.格—而影響良率。其所遭遇之問題包括有· ； 1·微影光罩（Photo _]〇之分配’若於^影製程中對機台 作光罩置換的動作，將增加非必要的操作時間與人力； 2. 機台產能（Ibd Capaeity) _，必_免機台處理數量 過多或過少； 3. 限定機台或指定相同步進機之限制； 4. 製程中關鍵層次與非關鍵層次的比例。 因此，對欲進行微影步驟的每一批次（L〇t)晶片而言， 選擇精確而適當的微影機台，除需以均衡機台之載荷量為前 提外，亦需保持工作效率^ 為此，本發明係揭露一種微影機台載荷分配系統之最佳 化方法，在各批次晶片進入各微影機台之前，係使用一套特 殊之派工演算法（Dispatching Algorithm) ’以製造執行系統 (Manufacturing Execution System，MES)以及電腦整合製 造系統（Computer Integrated Mamifacturing，CIM)加以整 合處理，不但可分配最適合的微影機台載荷量，以使適當之 機台分配於各次欲處理之晶片’亦可增加工作效率並且使 各機台之工作載荷達到最佳平衡狀態。 發明概述： 本發明之主要目的係提供一種微影機台載荷分配 (allocation)系統之最佳化方法，在各批次晶片進入各微影 3 ---- ----- ---聚 i I (請先閱讀背面之注意事項再填寫本頁) JSJ· 線- 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公餐） A7 46495〇 ----------2Z---- 五、發明說明（夕） 機台之前’先實行演算整合，再進行處理，使之分配最適合 的微影機台載荷量，以使適當之機台分配於各批次欲處理之 曰曰片亦可增加工作效率並且使各機台之處理達到最佳平衡。 為達上述之目的，本發明係提供一種微影機台载荷分配 系統之最佳化方法’在各批次晶#進人各微雜台製程動作 之則係使用一套特殊之派工演算法（Dispatching Algorithm)’以製造執行系統以及電腦整合製造系 統（CIM)加以整合處理’本剌之—較佳實關係包含下 列步驟： (a) 先計异出各欲進行製程之群組微影機台中，各該微影 機台於目前線上的工作載荷量（Current L〇ading ); (b) 將欲進行製程而尚未分派之各批次（L〇t)晶片，依 其產品型號（ProductID)分類； (c) 檢驗各初始製程路線中之各該批次晶片，是否有指定 '分派於單一特定機台之限定； (cl)將有限定分派於單一特定機台之批次晶片，直接在 製造執行系統（MES)標明（HighIight)該批次晶片所選 定之機台； (d) 檢驗各該批次晶片則檢驗該產品型號（jp)是否列入 機台限制表格中， (e) 將該機台限制表所列之機台作為該批次晶片之加工候 選機台； (el )未列入機台限制表格中的批次晶片，則所述群組微 4 i紙張尺度適用中國國家標準（CNS)aJ規格（2l^97公爱·) * 广 >. ϋ ί n n Jm i ϋ i * I I (請先閲讀背面之注意事項再填寫本頁) 訂· 464950 五、發明說明（f) 影機台皆可作為該批次晶片之加工候選機台； (f) 判斷各該等待派工之批次晶片型號與是否與先前已派 工予各該機台之批次晶片的型號相同； (fi)將各該等待派工之晶片型號與先前已派工予各該機 台之批次晶片的型號不相同之各該批次晶片，分派於所述 群組微影機台中，線上工作载荷量最少（Lightestbading) 之機台。 (g) 將該所屬型號相同之各該批次晶片，進行試算檢驗各 該等待派工批次晶片選定之機台，於再加入各該等待派工 批次晶片後的載荷量（Loading)總數是否在合理範圍内； (gl)各該等待派工之批次晶片，於該選定之機台再加入 各該荨待派工批次晶片後的载荷量（L〇ading)總數若超過 合理乾園，則將各該等待派工之批次晶片分派於所述群組 微'5V機σ中線上工作載荷量最少lading )之 機台，否則，則分派於先前已派工加工同晶片型號之機台。 (h) 將各該批次晶片分派至所選定之機台，並於製造執 行系統（MES)標明（Highlight)該批次晶片所選定之機 台。 (hi )再次汁异各該批次晶片所選定機台之線上工作載荷 量’確定各該機台之線上載荷量均未超過合理範圍； (h2)確定各該進行分派之批次（L〇t)晶片均已分配於 最適之機台。 圖示之簡聲說明： 國 S?^NS)A4 規―。x (請先閱讀背面之注意事項再填寫本頁> 裝 M; 訂· 46495 Ο Α7 -------^——-__ 五、發明說明（^) 圖一係為本發明實施例之機台組（步進機或掃瞄機）分 配系統。 圖二為實施本發明微影機台載荷分配系統之最佳化方法 , 的流程圖。 圖號說明： Κ)-批次晶片（Lot) 20, 30, 40-微影機台 20a，30a，40a-微影機台 50, 60, 70-微影機台 100, 110,120, 130, 140,150,160, 170, 180, 190, 200, 210, 230, 240-流程步驟 發明之詳細說明： 圖一所示為機台組分配系統，其中之機台可為I-Line步 進機或為深紫外光（DUV)步進機或掃瞄機之一種，以實施 本發明微影機台載荷分配系統之最佳化方法，其中機台2〇、 3〇、40皆屬同製程功能，所使用亦為同格式光罩，只有其製 矛玉特性具微小差異性（係如考慮加工效率或製程限制等因 素），而20a、30a、40a係為與上述機台20、30、40相搭配 (Match)之機台；而機台50、60、70則為處理精密度要求 較低之晶片層次（layer)製程。 請參閱圖二，係為實施本發明微影機台載荷分配系統之 最佳化方法的動作流程圖’因批次晶片進入微影機台之時間 為14小時’所以本發明之微影機台載荷分配系統之最佳化方 本紙❼度_ 準（CNS)A4規格（21〇^297公釐) -- (請先閱讀背面之注意事項再填寫本頁) 裝 翼 11 訂 --綠“ 46 4 9 j464950 A7 ----- B7 V. Description of the invention (2) For the semiconductor industry with extremely high suspension requirements, it is a problem that cannot be ignored. In addition, if you choose to change the designated machine during the lithography process, it will cause the system to overrun the rule. It will affect the yield. The problems encountered include: 1. The distribution of the photomask (Photo _) 〇 'If the photomask is replaced in the process of film production, it will increase unnecessary operation time and manpower; 2 . Machine capacity (Ibd Capaeity) _, must _ free machine too much or too little; 3. Limit the machine or specify the same stepper restrictions; 4. The proportion of critical and non-critical levels in the process. Therefore, For each batch (Lot) of wafers to be subjected to the lithography step, choose an accurate and appropriate lithography machine. In addition to the premise of equalizing the load of the machine, it is also necessary to maintain working efficiency. Therefore, the present invention discloses an optimization method for a lithography machine load distribution system. Before each batch of wafers enters each lithography machine, a special Dispatching Algorithm is used to manufacture The execution system (Manufacturing Execution System, MES) and computer integrated manufacturing system (Computer Integrated Mamifacturing, CIM) are integrated and processed. Not only can the most suitable lithography machine load be allocated, so that the appropriate machine is allocated to each The wafer to be processed can also increase the working efficiency and achieve the optimal balance of the work load of each machine. Summary of the invention: The main purpose of the present invention is to provide a method for optimizing the lithography machine load allocation system. In each batch of wafers enter each lithography 3 ---- ----- --- Poly i I (Please read the precautions on the back before filling out this page) JSJ · Line-This paper size applies to Chinese national standards (CNS) A4 specifications (210 X 297 meals) A7 46495〇 --------- 2Z ---- 5. Description of the invention (Even) Before the machine, perform calculation integration before processing. Making it allocate the most suitable lithography machine load, so that the appropriate machine can be distributed among the batches of films to be processed can also increase the work efficiency and achieve the best balance of the processing of each machine. For the purpose, the present invention is to provide a method for optimizing the load distribution system of lithography machines. When a batch of crystals enters each micro-hybrid process, a special dispatch algorithm is used (Dispatching Algorithm). ) 'To Manufacturing Execution System and Computer Integrated Manufacturing System (CIM) "Integrate processing"-the best real relationship includes the following steps: (a) First calculate the group of lithography machines that are different from each other to be processed, and the workload of each lithography machine on the current line (Current L〇ading); (b) batches (Lot) of wafers that have not yet been assigned for processing, are classified according to their product models (ProductID); (c) inspection of each batch of wafers in each initial process route , Is there a limit to specify 'assignment to a single specific machine; (cl) A batch of wafers that are limited to a single specific machine will be directly marked on the Manufacturing Execution System (MES) (HighIight) for the batch of wafers selected. Machines; (d) Inspection of each batch of wafers is to check whether the product model (jp) is listed in the machine restriction table, (e) The machine listed in the machine restriction table is used as the processing of the wafers of the batch Candidate machines; (el) Batch wafers that are not listed in the machine restriction table, the group micro 4 i paper size applies the Chinese National Standard (CNS) aJ specification (2l ^ 97 public love ·) * 广 >. ϋ ί nn Jm i ϋ i * II (Please read the notes on the back before filling this page) Order · 464950 V. Description of the invention (f) The camera can be used as a candidate processing machine for the batch of wafers; (f) Judge the model of each batch of wafers waiting to be dispatched and whether they are the same as those previously dispatched to each. The batches of wafers on the machine have the same model; (fi) The batches of wafers that are different from the batches of wafers that have been previously dispatched to each of the machines are assigned to the lot. Among the group lithography machines mentioned above, the machine with the least online workload (Lightestbading). (g) Perform a trial calculation on each batch of wafers of the same model to check the machine selected for each batch of wafers waiting for dispatch, and add the total amount of loading after adding each batch of wafers waiting for dispatch. Whether it is within a reasonable range; (gl) For each batch of wafers waiting to be dispatched, the total load (L0ading) after adding each of the batches of wafers waiting to be dispatched to the selected machine if the total amount exceeds Park, all batches of wafers waiting for dispatch will be assigned to the machine with the least amount of working load on the line of the group's 5V machine σ; otherwise, it will be assigned to the one that has previously been dispatched to process the same wafer model Machine. (h) Allocate each batch of wafers to the selected machine, and mark (Highlight) the machine selected by the batch in the Manufacturing Execution System (MES). (hi) Differentiating the online working load of the machine selected by each batch of wafers' again to confirm that the online load of each machine does not exceed a reasonable range; (h2) determining each batch to be allocated (L. t) The chips have been allocated to the most suitable machine. Brief description of the picture: National S? ^ NS) A4 Regulations. x (Please read the precautions on the back before filling in this page> Order M; order 46495 Α7 ------- ^ ——-__ 5. Description of the invention (^) Figure 1 is an embodiment of the present invention Machine group (stepper or scanner) distribution system. Figure 2 is a flow chart of the optimization method for implementing the lithography machine load distribution system of the present invention. Flowchart description: Κ)-batch of wafers ( Lot) 20, 30, 40-lithographic stage 20a, 30a, 40a-lithographic stage 50, 60, 70-lithographic stage 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 230, 240- Detailed description of the process step invention: Figure 1 shows a machine group distribution system, where the machine can be an I-Line stepper or a deep ultraviolet (DUV) stepper or a scanner. A method for optimizing the load distribution system of the lithography machine of the present invention, in which the machines 20, 30, and 40 all have the same process function, and the same format photomask is used, only the spear jade characteristics There are slight differences (such as consideration of processing efficiency or process limitations), and 20a, 30a, 40a are machines that match (Match) with the above machines 20, 30, 40; and machines 60, 70 was lower precision requirements of the wafer processing level (Layer) process. Please refer to FIG. 2, which is an operation flowchart for implementing the optimization method of the lithography machine load distribution system of the present invention, 'because the batch wafer enters the lithography machine time is 14 hours', so the lithography machine of the present invention Optimized Load Distribution System Paper Size _ Standard (CNS) A4 Specification (21〇 ^ 297mm)-(Please read the precautions on the back before filling this page) Wing 11 Order-Green "46 4 9 j
五、發明說明（厶〉 法每12小鳴作-次，其包含有下列步驟： 步驟（100):在半導體製程中的微影步驟，在各批次晶 片進入各微雜台之前’係讀料先計a每—微影機台 線上的载荷情形100，其計算方式如下：V. Description of the Invention (厶) The method consists of 12 small work-times, which includes the following steps: Step (100): The lithography step in the semiconductor process, before each batch of wafers enters each micro-hybrid stage First calculate the load situation a on each lithography machine line 100, and its calculation method is as follows:
Li ΣLi Σ
其中 a 機台i的線上載荷量（CurremL〇ading)， Ci=機台i的所能载荷的總量（Capacity)， =批次j所包含之晶片數量，該晶片品質要求所適 合之加工機台為機台i， Dij -批次j晶片於機台i所剩餘未加工之層數 (Layers), Tijs批次j晶片於機台丨加工至最後一層（LagtLayer) 所剩餘之時間，且刊係以理論上時間X批次之優先權重。 步驟（110):以該數學式計算出欲進行製程之群组微影 機台中，每一微影機台目前線上的工作載荷量之後（Current Loadmg)，將欲進行製程而尚未分派之各批次（L〇〇晶片， 依其產品型號（ProductID)分類。 進行此步驟（11〇) ’在後續分派步驟即可依序分派，避 免將Π產型號之批次分派到太多機台，而增加光罩 (Mask)置換的頻率，甚而減低操作效率。 步驟（120):檢驗各初始製程路線中（Original Route) ’ ----- -------.Γ'裝 i ί請先閱讀背面之注$項再填寫本頁> 訂Where a is the online load of machine i (CurremL〇ading), Ci = the total load of machine i (Capacity), = the number of wafers contained in batch j, the wafer quality requirements for the processing machine The machine is machine i, Dij-the number of unprocessed layers remaining on lot i of wafer j on machine i (Layers), the time remaining for Tijs lot j wafers to be processed on the machine to the last layer (LagtLayer), and published Based on the theoretical priority of time X batch. Step (110): Using the mathematical formula to calculate the group lithography machines to be processed, after each lithography machine is currently on-line with the workload (Current Loadmg), the batches to be processed but not yet assigned (L00 chip, sorted by its product model (ProductID). Perform this step (11)) in the subsequent distribution step can be sequentially distributed to avoid batches of Π production models to too many machines, and Increasing the frequency of Mask replacement, even reducing the operation efficiency. Step (120): Check the original route '----- -------. Γ' 装 i ί Please Please read the note $ on the back before filling in this page.
_;L 本紙張尺錢財關家標準（CNS)A_4規格咖公$7"_; L Paper Ruler Money and Family Standard (CNS) A_4 Specification Coffee $ 7 "
五、發明說明 ) 各該批次晶片是否已有限定只能分派於單一特定機台。 步驟（210) ·右批次晶片已限定只能分派於單一特定機 台之情形，則可以直接在製造執行系統（MES)檩明 (Highlight)該批次晶片所選定之機台，而不必再接受後續 之機台分派。 步驟（130):未限定機台之各該批次晶片，則檢驗該產 品型號（ID )是否列入機台限制表格（Tool constraint table ) 中。 此步驟（130)係檢查所述批次晶片所屬的產品型號是否 列在機台限制表格中，亦即檢查該型號是否有因為考慮加工 效率或製程限制等因素’而限制只能於某些機台上進行製程。 步驟（140 ):於上述步驟（130 )之檢驗，若產品型號（ID ) 確定列入機台限制表格中，則由該機台限制表所列之機台作 為該批次晶片之加工候選機台·。 步驟（150):於上述步驟（130)之檢驗，當產品型號（ID) 未列入機台限制表格中的批次晶片，則群組之同類型微影機 台皆可作為該批次晶片之加工候選機台。 步驟（160):再次檢驗確定各該批次晶#所選定之加工 機台是否有唯一之限定。 步驟（210):於上述步驟（16〇)之檢驗，若檢驗碟定只 可選定單一機台加工’直接在製造執行系統（MES)標明 (Highlight)該批次晶片所選定之機台，而不必再接受後續 之機台分派。 ________ 8 本紙張尺度適財國國家標準（CNS)A4規格<21〇 X 297公愛） 4 6 y.) A7 围· R7 五、發明說明（/) 步驟（170) ·於上述步驟（16〇)之檢驗’確定未限定單 -機台加卫之等待紅批次晶片，辭清各料待派工之批 次晶片型號與是否與先前已派工完畢之批次晶片型號相同。 該步驟（no)亦即判斷是否使用同一片光罩；此步驟主 要考慮黃光加工步驟需配合光罩使用之特性，亦即黃光機台 要加工-批次晶4，必須先將該批次晶片於曝光步驟所欲使 用之光罩置於機台上’且曝光未完成前不可取至其他機台使 用。若將使用同-片光罩的批次產品指定到不同的機台時， 由於-次只有]機使用刻光罩，因此容易造成多台機台 搶光罩，甚或造成機台停工等待使用光罩的情況外，同時於 不同機台間置換光罩也會造成加工效率（Efficiency)降低。 由於同型躺產品在相jgj的加工步_吏關—#光罩，因此 在不k成各機台的工作载荷（ToolLoading)差異太大的情況 下’盡量指定同-台機台加工，以減少因光罩置換而造成不 必要的產能浪費。 步驟（180):若上述步驟（170)同型號之各該批次晶片， 則進行試算各該等待派工批次晶片選定之機台於再加入各該 專待派工批次晶片後的載荷量（L〇ading)總數，撿驗是否在 合理範圍内。 步驟（190):若上述步驟（170)不同型號之各該批次晶 片’則分派於線上載荷量最少（LightestLoading)之機台。 步驟（210).接續步驟（190)直接在製造執行系就（mes) 標明（Highlight)該批次晶片所選定之機台。 _______ 9 本紙張國國家標準（CNS)A4i格(210 X 297公釐） c靖先聞讀背面之注意事項再填寫本頁)V. Description of the invention) Whether each batch of wafers is limited can only be assigned to a single specific machine. Step (210) · In the case where the right batch of wafers has been limited to a single specific machine, it can be directly selected in the manufacturing execution system (MES) to highlight the machine selected by the batch of wafers, without having to Accept subsequent machine assignments. Step (130): Each batch of wafers of the machine is not limited, then it is checked whether the product model (ID) is listed in the tool constraint table. This step (130) is to check whether the product model to which the batch of wafers belongs is listed in the machine restriction table, that is, to check whether the model is restricted to certain machines due to factors such as processing efficiency or process limitations. Process on stage. Step (140): In the inspection of the above step (130), if the product model (ID) is determined to be listed in the machine restriction table, the machine listed in the machine restriction table is used as a candidate machine for the batch of wafers. station·. Step (150): In the inspection of the above step (130), when the product model (ID) is not included in the batch of wafers in the machine restriction table, the lithographic machines of the same type in the group can be used as the batch of wafers. Processing candidate machine. Step (160): Check again to determine if the processing machine selected by each batch of crystals has a unique limitation. Step (210): In the inspection of the above step (16), if the inspection plate can only be selected for processing by a single machine, the machine selected by the batch of wafers will be marked (Highlight) directly in the manufacturing execution system (MES), and It is no longer necessary to accept subsequent machine assignments. ________ 8 The size of this paper is the National Standard (CNS) A4 Specification < 21〇X 297 Public Love) 4 6 y.) A7 Wai · R7 V. Description of the Invention (/) Step (170) 〇) Inspection 'to determine the unrestricted single-machine Jiawei waiting for the red batch of wafers, and clarify whether the batch wafer model of each material to be dispatched is the same as the batch wafer model that has been dispatched previously. This step (no) is to judge whether to use the same photomask; this step mainly considers the characteristics of the yellow light processing step that needs to be used with the photomask, that is, the yellow light machine is to process-batch of crystal 4, the batch must be first The photomask to be used in the exposure step is placed on the machine 'and cannot be taken to other machines until the exposure is completed. If a batch of products using the same photomask is assigned to different machines, since the machine is only used once, it is easy to cause multiple machines to grab the photomask, or even cause the machine to stop and wait for the use of light. In addition to the case of the mask, the replacement of the mask between different machines at the same time will also reduce the processing efficiency. Because the same type of lay products are processed in the same jgj step_LER 关 — # photomask, if the working load (ToolLoading) of each machine is not too different, 'designate the same machine as much as possible to reduce Unnecessary waste of production capacity due to mask replacement. Step (180): If the batch of wafers of the same model in the above step (170), perform a trial calculation of the load of each machine selected by the batch of wafers waiting for dispatch after adding the wafers of the batch to be dispatched exclusively. The total amount (L0ading), check whether it is within a reasonable range. Step (190): If the batches of wafers of different models in step (170) above are allotted to the machine with the least online load (Lightest Loading). Step (210). Following step (190), the machine selected for the batch of wafers is marked (Highlight) directly in the manufacturing execution system. _______ 9 National Standard (CNS) A4i of this paper (210 X 297 mm) c Jing first read the precautions on the back before filling out this page)
五、發明說明（I) 步驟（200):上述步驟（180)於加入等待派工批次晶片 後之載荷量（Loading)總數在合理範圍内之候選機台，則可 將各該批次晶片直接分派於候選之機台。 步驟(210 );接續步驟(2〇0 )直接在製造執行系統(MES ) 標明（Highlight)該批次晶片所選定之機台。 步驟（190):上述步驟（200)於加入等待派工批次晶片 後之载荷量（Loading)總數已超過合理範圍内之候選機台， 則分派於線上載荷量最少（LightestLoading)之機台。 步驟(210):接續步驊(190)直接在製造執行系統(MES) 標明（Highlight)該批次晶片所選定之機台。 步驟（220):最後再次計算各該批次晶片所選定機台之 線上載荷量。 步驟（230):確定所有批次之晶片都已分派好選定之機 ‘. 台。 步驟（110):於上述步驟（230)檢查出未被分派候選機 台之批次晶片重新分類。 步驟（240):確定所有批次之晶片都已分派之後，則開 始進行製程加工。 綜上所述，本發明微影機台載荷分配系統之最佳化方法 係在各批次晶片進入各微影機台之則’係使用一套特殊之派 工演算法（Dispatching Algorithm)’以製造執行系統（MES ) 以及電腦整合製造系統（CIM)加以整合處理，其係具有下 列優點： 10 本纸張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） -11111111111 — ΙΓ - I I (請先閲讀背面之注項再填寫本頁) 訂. JI ‘ .線· 46495 0 B7 五、發明說明（|-) 1. 應用本發明可使得深紫外光（DUV)步進機或掃瞄機與其 搭配之I-Line微影機台製程更穩定，並且使其載荷量達到 平衡。 2. 減少機台閒置時間（L〇st Time)。 3. 消弭疊對超出規格而須重加工之比例，減少產能浪費，並 尚產品良率。 4. 明顯減少更換操作機台的頻率。 此外，本發明微影機台載荷分配系統之最佳化方法，完 全適用目前微影機台載荷分配系統，不僅檢測過程不會難以 實行’且更不需提高製程技術，並能克服習用技術之問題。 由此可知，本發明之實用性及進步性顯應具備β此外，本發 明於申清刖並未曾見於任何公開場合或刊物上，因此本案深 具「實用性、新穎性及進步性」之發明專利要件，故爰法提 出發明專利之申請。祁請貴審查委員允撥時間惠允審查並早 賜與專利為禱。 以上所述係利用較佳之實施例來詳細說明本發明，而本 發明涵蓋之範圍並不限於其所示之實施例，而且熟知該技藝 的人士皆能明瞭，適當而作些微之改變及調整，仍將不失本 發明之要義所在’亦不脫離本發明之精神與範園，故都應視 為本發明之進一步實施狀況。 ____ . 11 本紙張ϋ適用中國國家標準"(CNS)A4規格(210 X 297公楚） (請先閱讀背面之注意事項再填寫本頁) b裝.V. Description of the invention (I) Step (200): In the above step (180), after adding a candidate machine whose total load is within a reasonable range after adding batches of wafers waiting to be dispatched, each batch of wafers may be Directly assigned to candidate machines. Step (210); The subsequent step (200) directly marks (Highlight) the machine selected by the batch of wafers in the manufacturing execution system (MES). Step (190): In the above step (200), after adding a batch of wafers waiting to be dispatched to a lot of candidates, the number of loads that has exceeded a reasonable range is assigned to the machine with the smallest online load (Lightest Loading). Step (210): Following step (190), the machine selected by the batch of wafers is marked (Highlight) directly on the manufacturing execution system (MES). Step (220): Finally calculate the line load of the machine selected by each batch of wafers. Step (230): Make sure that all batches of wafers have been assigned the selected machine ′. Step (110): In step (230) above, it is checked that the batch of wafers that have not been assigned a candidate machine is reclassified. Step (240): After confirming that all batches of wafers have been dispatched, the process begins. In summary, the optimization method of the lithography machine load distribution system of the present invention is to use a special set of Dispatching Algorithm when each batch of wafers enters each lithography machine. The manufacturing execution system (MES) and computer integrated manufacturing system (CIM) are integrated and processed, which has the following advantages: 10 This paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 mm) -11111111111 — ΙΓ- II (Please read the note on the back before filling this page) Order. JI '. Line · 46495 0 B7 V. Description of the invention (|-) 1. The application of the present invention can make a deep ultraviolet (DUV) stepper or scanner The sighting machine and its matching I-Line lithography machine have a more stable manufacturing process and make their load balance. 2. Reduce the machine idle time (Lost Time). 3. Eliminate the proportion of over-specification that requires heavy processing, reduce waste of production capacity, and maintain product yield. 4. Significantly reduces the frequency of changing the operating platform. In addition, the method for optimizing the lithography machine load distribution system of the present invention is completely applicable to the current lithography machine load distribution system, not only will the detection process not be difficult to implement, but also does not need to improve the process technology, and can overcome the conventional technology problem. It can be seen that the practicality and progressiveness of the present invention should obviously have β. In addition, the present invention has not been seen in any public places or publications in Shen Qingyu, so this case is deeply "practical, novel, and progressive". Patent requirements, therefore, the law filed an application for an invention patent. Qi asked your examiner to allow time for examination and to grant the patent early. The above is a detailed description of the present invention using the preferred embodiments, and the scope of the present invention is not limited to the embodiments shown, and those skilled in the art will understand that appropriate changes and adjustments will be made. The essence of the present invention will not be lost, nor will it depart from the spirit and scope of the present invention, so it should be regarded as the further implementation status of the present invention. ____. 11 This paper is compliant with the Chinese National Standard " (CNS) A4 Specification (210 X 297 Gongchu) (Please read the precautions on the back before filling this page) b Pack.