TWI717269B - Memory unit operation method - Google Patents
Memory unit operation method Download PDFInfo
- Publication number
- TWI717269B TWI717269B TW109113475A TW109113475A TWI717269B TW I717269 B TWI717269 B TW I717269B TW 109113475 A TW109113475 A TW 109113475A TW 109113475 A TW109113475 A TW 109113475A TW I717269 B TWI717269 B TW I717269B
- Authority
- TW
- Taiwan
- Prior art keywords
- data writing
- switching element
- voltage
- resistance element
- memory cell
- Prior art date
Links
Images
Landscapes
- Semiconductor Memories (AREA)
Abstract
Description
本揭示內容關於一種記憶單元操作方法,特別是能在記憶單元內設定不同阻抗值,以儲存不同邏輯訊號之技術。The present disclosure relates to a method for operating a memory cell, particularly a technology capable of setting different impedance values in the memory cell to store different logic signals.
隨著電腦速度越來越快,對於記憶體的速度以及穩定性的要求越來越高。近年來,對於非揮發性記憶體的設計的相關討論越來越多。非揮發性記憶體有多種實施類型,例如包含相變化記憶體、電阻式記憶體、磁阻式記憶體等等。然而,目前上述的記憶體可能會因為一些製程(例如為回焊製程等)的操作溫度過高,而造成記憶體內部的資料發生錯誤。As computers get faster and faster, the requirements for memory speed and stability are getting higher and higher. In recent years, there have been more and more discussions on the design of non-volatile memory. There are many types of non-volatile memory, including phase change memory, resistive memory, magnetoresistive memory, and so on. However, the above-mentioned memory may cause errors in the internal data of the memory due to the high operating temperature of some processes (for example, reflow process, etc.).
本揭示內容係關於一種記憶單元操作方法,包含下列步驟:在資料寫入期間,導通記憶單元之第一開關元件,以從與記憶單元電性連接之匹配線接收第一資料寫入電壓,使得與第一開關元件串聯連接之第一電阻元件具有對應於第一資料寫入電壓之第一阻抗值。在資料寫入期間,關斷記憶單元之第二開關元件,使得與第二開關元件串聯連接之第二電阻元件與匹配線之間保持斷路並具有第二阻抗值。第二阻抗值大於第一阻抗值。The present disclosure relates to a method for operating a memory cell, which includes the following steps: during a data writing period, a first switching element of the memory cell is turned on to receive a first data writing voltage from a match line electrically connected to the memory cell, so that The first resistance element connected in series with the first switching element has a first impedance value corresponding to the first data writing voltage. During the data writing period, the second switching element of the memory cell is turned off, so that the second resistance element connected in series with the second switching element and the matching line remain open and have a second impedance value. The second impedance value is greater than the first impedance value.
本揭示內容還關於一種記憶單元操作方法,包含下列步驟:在資料寫入期間,透過匹配線對記憶單元中之第一電阻元件施加第一資料寫入電壓,使第一電阻元件具有對應於第一資料寫入電壓之第一阻抗值。其中匹配線係電性連接於第一電阻元件。在資料寫入期間,維持記憶單元之第二電阻元件與匹配線間之斷路,使第二電阻元件在無電流通過的情況下具有第二阻抗值。匹配線係電性連接於第二電阻元件。第二阻抗值大於該第一阻抗值。The present disclosure also relates to a method for operating a memory cell, including the following steps: during a data writing period, applying a first data writing voltage to a first resistance element in the memory cell through a matching line, so that the first resistance element has a A first impedance value of the data writing voltage. The matching line is electrically connected to the first resistance element. During the data writing period, the disconnection between the second resistance element of the memory cell and the matching line is maintained, so that the second resistance element has a second impedance value when no current flows. The matching line is electrically connected to the second resistance element. The second impedance value is greater than the first impedance value.
據此,由於第二阻抗值遠大於第一阻抗值,故當後續執行資料搜索時,匹配搜索結果的記憶單元將能維持高阻抗,以維持匹配線上的電壓。此外,不匹配搜索結果的記憶單元則具有低電壓,使匹配線上的電壓產生明顯的變化,以提昇搜索精確性。Accordingly, since the second impedance value is much greater than the first impedance value, when the data search is subsequently performed, the memory cell matching the search result will be able to maintain a high impedance to maintain the voltage on the matching line. In addition, the memory cell that does not match the search result has a low voltage, which causes the voltage on the matching line to change significantly to improve the search accuracy.
以下將以圖式揭露本發明之複數個實施方式,為明確說明起見,許多實務上的細節將在以下敘述中一併說明。然而,應瞭解到,這些實務上的細節不應用以限制本發明。也就是說,在本發明部分實施方式中,這些實務上的細節是非必要的。此外,為簡化圖式起見,一些習知慣用的結構與元件在圖式中將以簡單示意的方式繪示之。Hereinafter, multiple embodiments of the present invention will be disclosed in the form of drawings. For clear description, many practical details will be described in the following description. However, it should be understood that these practical details should not be used to limit the present invention. That is, in some embodiments of the present invention, these practical details are unnecessary. In addition, in order to simplify the drawings, some conventionally used structures and elements will be shown in a simple schematic manner in the drawings.
於本文中,當一元件被稱為「連接」或「耦接」時,可指「電性連接」或「電性耦接」。「連接」或「耦接」亦可用以表示二或多個元件間相互搭配操作或互動。此外,雖然本文中使用「第一」、「第二」、…等用語描述不同元件,該用語僅是用以區別以相同技術用語描述的元件或操作。除非上下文清楚指明,否則該用語並非特別指稱或暗示次序或順位,亦非用以限定本發明。In this text, when a component is referred to as “connected” or “coupled”, it can be referred to as “electrically connected” or “electrically coupled”. "Connected" or "coupled" can also be used to mean that two or more components cooperate or interact with each other. In addition, although terms such as “first”, “second”, etc. are used herein to describe different elements, the terms are only used to distinguish elements or operations described in the same technical terms. Unless clearly indicated by the context, the terms do not specifically refer to or imply order or sequence, nor are they used to limit the present invention.
本揭示內容係關於一種記憶單元的操作方法。第1A係根據本揭示內容之部份實施例的記憶單元100(memory cell)的示意圖。如第1A圖所示,記憶單元100用以電性連接於匹配線ML及搜索線SL1、SL2,其中搜索線SL1、SL2的邏輯準位為互為相反(即,當搜索線SL1具有致能電壓的準位時,搜索線SL2具有禁能電壓的準位)。The present disclosure relates to an operating method of a memory unit. 1A is a schematic diagram of a memory cell 100 (memory cell) according to some embodiments of the present disclosure. As shown in FIG. 1A, the
第1B係根據本揭示內容之部份實施例的記憶體裝置200的示意圖。如第1B圖所示,記憶體裝置200包含多個如第1A圖所示的記憶單元100,且多條匹配線ML及搜索線SL1、SL2連接於相應的記憶單元100。在部份實施例中,上述多個記憶單元100係以陣列形式設置於記憶體裝置200中,以儲存各種不同的資料(如:路由表)。1B is a schematic diagram of the
在一實施例中,當記憶體裝置200執行資料寫入(write)之動作時(以下簡稱「資料寫入期間」),匹配線ML用以被充電至特定的寫入電壓,以調整記憶單元100內的阻抗特性。不同的阻抗特性將對應於不同的邏輯訊號。在一實施例中,當記憶體裝置200執行資料搜索(search)之動作時(以下簡稱「資料搜索期間」),搜索線SL1、SL2提供對應的禁能或致能電壓至記憶單元100,以控制記憶單元100內不同路徑的導通或關斷。據此,根據匹配線ML上的電壓變化,記憶體裝置200即可判斷記憶單元100內儲存之邏輯訊號是否與搜索指令中欲搜索的資訊(以下簡稱「搜索目標」)相符。In one embodiment, when the
第2A圖係根據本揭示內容之部份實施例的記憶單元100之等效電路示意圖。在一實施例中,記憶單元100為三態內容可定址記憶體(Ternary Content Addressable Memory,TCAM),可運用於網路路由器中,但本揭示內容並不以此為限。由於本領域人士能理解三態內容可定址記憶體及記憶體裝置的結構及原理,故在此不另贅述。FIG. 2A is a schematic diagram of an equivalent circuit of the
如第2A圖所示,記憶單元100包含第一電路110及第二電路120。第一電路110及第二電路120係電性連接於匹配線ML與字元線BL之間。在一實施例中,字元線BL係連接於參考電位(如:接地電壓)。第一電路110包含第一開關元件T1及第一電阻元件R1。第二電路120包含第二開關元件T2及第二電阻元件R2。匹配線ML分別電性連接於第一電阻元件R1之一端及第二電阻元件R2之一端。第一開關元件T1與第一電阻元件R1相串聯,第二開關元件T2與第二電阻元件R2相串聯。第一電路110與第二電路120則相並聯。As shown in FIG. 2A, the
第2B圖係根據本揭示內容之部份實施例的開關元件之半導體結構的示意圖。在一實施例中,第2B圖所示之半導體結構可用以實現第2A圖所示之第一開關元件T1及/或第二開關元件T2,但本案不以此為限。如第2B圖所示,開關元件之半導體結構包含閘極Tg、汲極Td及源極Ts。當閘極Tg接收到致能電壓時,將能於汲極Td及源極Ts間形成導通路徑,根據汲極Td及源極Ts間的電壓差形成不同大小的導通電流。FIG. 2B is a schematic diagram of the semiconductor structure of the switching element according to some embodiments of the present disclosure. In one embodiment, the semiconductor structure shown in FIG. 2B can be used to realize the first switching element T1 and/or the second switching element T2 shown in FIG. 2A, but this case is not limited to this. As shown in FIG. 2B, the semiconductor structure of the switching element includes a gate Tg, a drain Td, and a source Ts. When the gate Tg receives the enabling voltage, a conduction path can be formed between the drain Td and the source Ts, and conduction currents of different magnitudes are formed according to the voltage difference between the drain Td and the source Ts.
在不同實施例中,第2A圖所示之第一電阻元件R1亦可形成於第一開關元件T1中。舉例而言,第一電阻元件R1可由第一開關元件T1中具有特定阻抗值的半導體結構來實現(詳細結構將於後續段落中說明)。意即,第一電阻元件R1及第一開關元件T1可整合於同一個電晶體結構中。為便於說明,在第2A圖中將兩個元件分別繪示,第一開關元件T1視為理想開關元件、第一電阻元件R1則為第一電路110中的阻抗部份,且第一電阻元件R1的阻抗值將會根據第一電路110(或第一開關元件T1)的跨壓而改變。In different embodiments, the first resistance element R1 shown in FIG. 2A can also be formed in the first switching element T1. For example, the first resistance element R1 can be realized by a semiconductor structure having a specific impedance value in the first switching element T1 (the detailed structure will be described in subsequent paragraphs). That is, the first resistance element R1 and the first switching element T1 can be integrated in the same transistor structure. For ease of description, in Figure 2A, the two elements are shown separately, the first switching element T1 is regarded as an ideal switching element, the first resistance element R1 is the impedance part of the
相應於上述實施例,在不同實施例中,第2A圖所示之第二電阻元件R2亦可形成於第二開關元件T2中,且第二電阻元件R2的阻抗值會根據第二電路120(或第二開關元件T2)的跨壓而改變。Corresponding to the above embodiment, in different embodiments, the second resistance element R2 shown in Figure 2A can also be formed in the second switching element T2, and the impedance value of the second resistance element R2 will be based on the second circuit 120 ( Or the cross voltage of the second switching element T2) changes.
操作上,在資料寫入期間,匹配線ML被充電至第一資料寫入電壓,且第一電路110導通,使得第一電阻元件R1具有對應於第一資料寫入電壓的第一阻抗值。換言之,第一電阻元件R1的阻抗值會根據導通第一電路110的第一資料寫入電壓進行改變,由初始阻抗值改變為第一阻抗值。在部份實施例中,第一資料寫入電壓為相對低電壓。在一些實施例中,第一資料寫入電壓為1~4伏特,但本揭示內容並不以此為限,第一資料寫入電壓的數值可依據實際製作記憶體的材料、製程或其他參數而調整。In operation, during the data writing period, the match line ML is charged to the first data writing voltage, and the
在一些實施例中,當匹配線ML充電至第一資料寫入電壓時,第一開關元件T1被導通、第二開關元件T2被關斷,使得第一電路110上響應於導通之第一開關元件T1及第一寫入電壓而形成第一電流。In some embodiments, when the match line ML is charged to the first data writing voltage, the first switching element T1 is turned on and the second switching element T2 is turned off, so that the first switch on the
同時,在資料寫入期間及匹配線ML具有第一資料寫入電壓時,第二電阻元件R2與匹配線ML間係維持斷路狀態,使第二電路120不會因為第一資料寫入電壓而產生電流。此時,第二電阻元件R2在無電流流過的情況下具有第二阻抗值,且第二阻抗值將遠大於第一阻抗值。在部份實施例中,第二開關元件T2於資料寫入期間始終保持關斷,確保第二電路120上不會產生電流,且第二開關元件T2從未被電流貫穿。At the same time, during the data writing period and when the matching line ML has the first data writing voltage, the second resistance element R2 and the matching line ML maintain an open state, so that the
為便於說明,在此將第一電阻元件R1/第二電阻元件R2(或第一電路110/第二電路120)對應於第一資料寫入電壓而形成的第一阻抗值稱為「低阻抗狀態(Low Resistance State)」,且將第二阻抗值則稱為「初始阻抗狀態」。如前述實施例所述,當第一電路110為低阻抗狀態、第二電路為初始阻抗狀態時,可用以紀錄邏輯訊號「0」。相對地,若第一電路110為初始阻抗狀態、第二電路為低阻抗狀態時,可用以紀錄邏輯訊號「1」。若第一電路110、第二電路120皆為初始阻抗狀態,則對應於邏輯訊號「X(don’t care)」,代表搜索時可忽略。For ease of description, the first resistance value formed by the first resistance element R1/second resistance element R2 (or the
在記憶體裝置200進行資料搜索的動作時(以下簡稱「資料搜索期間」),搜索線SL1、SL2係根據搜索指令被控制於致能準位或禁能準位,以控制第一開關元件T1及第二開關元件T2的導通及關斷。透過判斷匹配線ML上的電壓變化,記憶體裝置200即可確認記憶單元100中儲存的邏輯訊號是否為搜索目標。在一實施例中,若記憶單元100中儲存的邏輯訊號確實為搜索目標,則第一電路110及第二電路120中的其中之一會因為搜索線(SL1或SL2)的禁能而保持斷路,第一電路110及第二電路120中的其中之另一個則會具有「初始阻抗狀態」的高阻抗特性,使得匹配線ML上幾乎無電壓變化。本揭示內容利用第二電阻元件R2(或第二開關元件T2中的半導體結構,將於後續段落詳述)未曾被電流貫穿時,第二電阻元件R2具有的高阻抗特性,確保匹配線ML於搜索到正確目標時,不會產生漏電壓,避免記憶體裝置200不必要的耗電。When the
另一方面,由於「第一阻抗值(低阻抗狀態)」及「第二阻抗值(初始阻抗狀態)」間具有足夠大的阻抗差異,因此,在資料搜索期間,匹配搜索目標的記憶單元100及不匹配搜索目標的記憶單元100之間,將會因為阻抗的明顯不同,使得匹配線ML上的電壓變化有極為明顯的差異,藉此能提昇記憶體裝置200於進行資料搜索時的精確性。On the other hand, since there is a sufficiently large impedance difference between the "first impedance value (low impedance state)" and the "second impedance value (initial impedance state)", during the data search period, the
在此說明記憶單元100之「資料寫入程序」如後。第3圖為資料寫入程序的步驟流程圖。下述為方便清楚說明第3圖起見,係以第2A圖所示實施例搭配第3圖進行說明,然而第3圖所示實施例的應用不以第2A圖的實施例為限。The "data writing procedure" of the
在步驟S301,記憶體裝置200中的處理器(圖中未示)判斷寫入指令中欲寫入之邏輯訊號的準位為「1」、「0」或者是「X(don’t care)」。在記憶體裝置200中的處理器判斷需寫入的邏輯訊號為「0」時,執行步驟S302,處理器將匹配線ML充電至第一資料寫入電壓。在步驟S303中,導通第一開關元件T1,使第一電阻元件R1具有對應於第一資料寫入電壓之第一阻抗值,且在第一電路110上形成「低阻抗狀態」。In step S301, the processor (not shown in the figure) in the
在步驟S304中,在資料寫入期間,維持關斷第二開關元件T2,使第二電阻元件R2與匹配線ML間保持斷路。此時第二電路120或第二電阻元件R2上未曾有電流產生。此外,由於第二電阻元件R2未曾被電流所貫穿而具有第二阻抗值,因此能使第二電路120形成「初始阻抗狀態」。In step S304, during the data writing period, the second switching element T2 is kept turned off, so that the second resistance element R2 and the match line ML are kept open. At this time, no current is generated on the
在記憶體裝置200中的處理器判斷需寫入的邏輯訊號為「1」時,執行步驟S305。處理器將匹配線ML充電至第一資料寫入電壓。在步驟S306中,導通第二開關元件T2,使第二電阻元件R2具有對應於第一資料寫入電壓之第一阻抗值,且在第二電路120上形成「低阻抗狀態」。When the processor in the
在步驟S307中,在資料寫入期間,維持關斷第一開關元件T1,使第一電阻元件R1與匹配線ML間保持斷路。此時第一電阻元件R1因為未曾被電流流經而具有第二阻抗值,且使第一電路110形成「初始阻抗狀態」。In step S307, during the data writing period, the first switching element T1 is kept turned off, so that the first resistance element R1 and the match line ML are kept open. At this time, the first resistance element R1 has the second impedance value because no current flows through it, and the
在判斷需寫入的邏輯訊號為「X」時,執行步驟S308,在資料寫入期間,維持關斷第一開關元件T1,使第一電阻元件R1與匹配線ML間保持斷路。此時第一電阻元件R1因為未曾被電流流經而具有第二阻抗值,且在第一電路110上形成「初始阻抗狀態」。When it is judged that the logic signal to be written is "X", step S308 is executed. During the data writing period, the first switching element T1 is kept off, so that the first resistance element R1 and the matching line ML are kept open. At this time, the first resistance element R1 has a second impedance value because no current flows through it, and an "initial impedance state" is formed on the
同樣地,在步驟S309中,關斷第二開關元件T2,使第二電阻元件R2與匹配線ML間保持斷路。此時第二電阻元件R2因為未曾被電流流經而具有第二阻抗值,且在第二電路120上形成「初始阻抗狀態」。Similarly, in step S309, the second switching element T2 is turned off, so that the second resistance element R2 and the matching line ML are kept open. At this time, the second resistance element R2 has a second impedance value because no current flows through it, and an "initial impedance state" is formed on the
在一實施例中,記憶單元100係利用其未曾被導通時具備高阻抗之「初始阻抗狀態」的特性,作為記憶邏輯訊號的特性。因此,記憶單元100應用於「一次性寫入」的記憶體時,可具有更佳的儲存效果。In one embodiment, the
請參閱第4圖,第4圖係對應於第一電路110/第二電路120的半導體結構的特性圖。其中,第4圖的橫軸為阻抗值、縱軸為出現對應之阻抗值的機率(在本實施例中,縱軸數值係經過多次測試後統計出的數量累積百分比)。根據不同的阻抗特性,開關元件具有「低阻抗狀態LRS」、「高阻抗狀態HRS」及「初始阻抗狀態ORS」。「低阻抗狀態LRS」為開關元件響應於第一資料寫入電壓而導通時,對應之電阻元件具有的第一阻抗值(如:10
3~10
5歐姆)。「高阻抗狀態HRS」為開關元件響應於第二資料寫入電壓而導通時,對應之電阻元件具有的第三阻抗值(如:10
5~10
8歐姆),其中第二資料寫入電壓相對於第一資料寫入電壓為高電壓。在部份實施例中,第二資料寫入電壓為-1~-2伏特,但本揭示內容並不以此為限,第二資料寫入電壓的數值的數值可依據實際製作記憶體的材料、製程或其他參數而調整。「初始阻抗狀態ORS」則為開關元件未曾被導通時,對應之電阻元件所呈現出的第二阻抗值(如:10
9~10
10歐姆)。根據製程的差異,第一阻抗值、第二阻抗值及第三阻抗值將不會是一個固定值,而是一個阻抗範圍。如第4圖所示,「高阻抗狀態HRS」對應之第三阻抗值係介於「低阻抗狀態LRS」對應之第一阻抗值及「初始阻抗狀態ORS」對應的第二阻抗值之間。「初始阻抗狀態ORS」具有最高的阻抗特性。
Please refer to FIG. 4, which is a characteristic diagram of the semiconductor structure corresponding to the
如第4圖所示之特性圖,曲線L1為開關元件響應於第一資料寫入電壓而第一次導通時,對應之電阻元件的阻抗特性。曲線H1為開關元件響應於第一資料寫入電壓導通後,再響應於第二資料寫入電壓而第二次導通時,對應之電阻元件的阻抗特性。由第4圖可知,當匹配線ML上的電壓被控制於第一資料寫入電壓、第二資料寫入電壓間切換,使記憶單元100中的電路反覆地處於「高阻抗狀態HRS」及「初始阻抗狀態ORS」時,阻抗特性也會隨之改變。曲線Ln為開關元件響應於第一資料寫入電壓而第N次導通時,對應之電阻元件呈現出的阻抗特性。As shown in the characteristic diagram in Fig. 4, the curve L1 is the impedance characteristic of the corresponding resistance element when the switching element is turned on for the first time in response to the first data writing voltage. The curve H1 is the impedance characteristic of the corresponding resistance element when the switch element is turned on in response to the first data writing voltage and then turned on for the second time in response to the second data writing voltage. It can be seen from Figure 4 that when the voltage on the match line ML is controlled to switch between the first data writing voltage and the second data writing voltage, the circuit in the
承上,如第4圖所示,曲線Hn為開關元件響應於第二資料寫入電壓而第N次導通時,對應之電阻元件呈現出的阻抗特性。經過多次的切換,曲線會逐漸朝阻抗更低的方向變化。在一實施例中,若第一開關元件T1需被控制於「低阻抗狀態」,則第一開關元件T1透過匹配線ML反覆接收第一資料寫入電壓及第二資料寫入電壓,直到第一開關元件T1接收第一資料寫入電壓時,第一電阻元件R1之阻抗值等於或低於阻抗門檻值(如:10
4歐姆)。換言之,第一電路110將被反覆控制於「高阻抗狀態HRS」及「低阻抗狀態LRS」間,具體方法為:在資料寫入期間,在第一開關元件T1響應於第一資料寫入電壓導通時,匹配線ML上之電壓被調整為第二資料寫入電壓,使第一開關元件T1接收第二資料寫入電壓。接著,匹配線ML上之電壓再次被調整回第一資料寫入電壓,使第一開關元件T1再次接收第一資料寫入電壓。依據前述步驟反覆執行N次。在一實施例中,N為大於8之正整數。意即,匹配線ML切換第一資料寫入電壓及第二資料寫入電壓的次數大於8次。第4圖所示之曲線Ln為開關元件被第10次響應於第一資料寫入電壓時,對應之電阻元件呈現的特性曲線。
In addition, as shown in Figure 4, the curve Hn is the impedance characteristic of the corresponding resistance element when the switching element is turned on for the Nth time in response to the second data writing voltage. After many times of switching, the curve will gradually change towards a lower impedance. In one embodiment, if the first switching element T1 needs to be controlled in the "low impedance state", the first switching element T1 repeatedly receives the first data writing voltage and the second data writing voltage through the match line ML until the first a first data switching element T1 receives the write voltage, the resistance value of the first resistor element R1 is equal to or lower than the threshold impedance (eg: 104 ohms). In other words, the
請參閱第5圖,第5圖為記憶單元100中第一電路110及/或第二電路120的開關元件的另一實施例的半導體結構示意圖。在一實施例中,前述開關元件包含閘極Tg、汲極Td及源極Ts。其中,閘極Ts還包含浮動閘極Tf,浮動閘極Tf係位於兩層絕緣層Ti之間。汲極Td及源極Ts上則設有金屬層Tm。此外,如前所述,在半導體結構中可同時形成有電阻元件。如第5圖所示,汲極Td上設有介電層Te。介電層Te具有特定阻抗值,可作為前述實施例中之第一電阻元件R1及/或第二電阻元件R2,或作為前述實施例中第一電阻元件R1及/或第二電阻元件R2之至少一部份的半導體結構。舉例來說,在第一電路110/第二電路120被施加第一資料寫入電壓或第二資料寫入電壓時,介電層Te將會被貫穿,而形成低阻抗狀態LRS。Please refer to FIG. 5, which is a schematic diagram of a semiconductor structure of another embodiment of the switching element of the
在此進一步舉例說明記憶單元100之「資料搜索程序」如後。如前述步驟,記憶單元100係根據下列特性紀錄邏輯訊號:
在一實施例中,根據搜索目標的不同,搜索線SL1、SL2將根據下列表格調整致能電壓或禁能電壓,以導通或關斷對應之開關元件:
如上述兩個表格所示,當資料搜索期間,在記憶體裝置200判斷搜索指令中之搜索目標為「0」的情形下,搜索線SL1將關斷第一開關元件T1、搜索線SL2則導通第二開關元件T2。若記憶單元100內儲存的邏輯訊號與搜索目標匹配(即,「0」),則第一電路110、第二電路120分別為低阻抗狀態LRS及初始阻抗狀態ORS。此時,雖然第二開關元件T2被導通,但由於第二電路120處於初始阻抗狀態ORS,具有極高電阻,因此匹配線ML上並不會產生明顯的電壓變化,而能使記憶體裝置200精確地搜索出儲存有搜索目標的記憶單元100。此外,由於在資料搜索期間,匹配線ML上的電壓值小於資料寫入期間充電的電壓(即,第一資料寫入電壓或第二資料寫入電壓),因此第二開關元件T2(如:第5圖中之介電層Te)不會因為被電流貫穿而破壞最初的阻抗特性,使得第二開關元件T2能維持在初始阻抗狀態ORS。As shown in the above two tables, during data search, when the
另一方面,若搜索指令之搜索目標為「0」,但記憶單元100內儲存的邏輯訊號與搜索目標並不匹配(即,「1」),此時,第二開關元件T2同樣會被導通,但第二電路120處於低阻抗狀態LRS,具有低電阻。因此,匹配線ML會產生明顯的電壓變化,使得記憶體單元200能輕易地分辨出「不匹配」之搜索結果。On the other hand, if the search target of the search command is "0", but the logic signal stored in the
請參閱第6圖,第6圖為本揭示內容之部份實施例中記憶單元的應用示意圖。記憶體裝置200包含多個記憶單元100,該些記憶單元100可組成多個記憶資料庫TCAM1~TCAM8。在一實施例中,記憶資料庫TCAM1~TCAM8用以儲存對應於神經網路的資料庫,分別儲存不同類型的資料。由於記憶單元100係作為資料庫,因此只須執行一次資料寫入程序。Please refer to FIG. 6, which is a schematic diagram of the application of the memory unit in some embodiments of the disclosure. The
舉例而言,當神經網路600接收到待比對的分析資料(如:一張動物的影像)時,分析資料將經過神經網路600的多層神經分析單元CV1~CV4,逐一判斷分析資料內的特徵向量,最後經過連接層Ly(connected layer)產生分析後的一組邏輯特徵向量LSH(binary vector)。邏輯特徵向量LSH可作為輸入至記憶體裝置200中的搜索指令,以在多種類別的記憶資料庫TCAM1~TCAM8(如:陸地動物、水生動物等不同資料庫)中搜索出對應的結果(如:兔子的影像)。由於本領域人士能理解神經網路的運作及分析原理,故在此不另贅述。For example, when the neural network 600 receives the analysis data to be compared (such as an image of an animal), the analysis data will pass through the multi-layer neural analysis units CV1 to CV4 of the neural network 600, and determine the analysis data one by one. Finally, through the connected layer Ly (connected layer), a set of logical feature vectors LSH (binary vector) is generated after analysis. The logical feature vector LSH can be used as a search command input to the
前述各實施例中的各項元件、方法步驟或技術特徵,係可相互結合,而不以本揭示內容中的文字描述順序或圖式呈現順序為限。The various elements, method steps, or technical features in the foregoing embodiments can be combined with each other, and are not limited to the order of description or presentation of figures in the present disclosure.
雖然本揭示內容已以實施方式揭露如上,然其並非用以限定本揭示內容,任何熟習此技藝者,在不脫離本揭示內容之精神和範圍內,當可作各種更動與潤飾,因此本揭示內容之保護範圍當視後附之申請專利範圍所界定者為準。Although the content of this disclosure has been disclosed in the above manner, it is not intended to limit the content of this disclosure. Anyone who is familiar with this technique can make various changes and modifications without departing from the spirit and scope of this disclosure. Therefore, this disclosure The scope of protection of the content shall be subject to the scope of the attached patent application.
100:記憶單元 110:第一電路 120:第二電路 200:記憶體裝置 600:神經網路 T1:第一開關元件 T2:第二開關元件 R1:第一電阻元件 R2:第二電阻元件 S301-S309:步驟 ML:匹配線 SL1:搜索線 BL:字元線 Ts:源極 Td:汲極 Tg:閘極 Tb:基極 Tf:浮動閘極 Ti:絕緣層 Te:介電層 Tm:金屬層 L1:曲線 Ln:曲線 H1:曲線 Hn:曲線 LRS:低阻抗狀態 HRS:高阻抗狀態 ORS:初始阻抗狀態 TCAM1-TCAM8:記憶資料庫 CV1-CV4:神經分析單元 Ly:連接層 LSH:邏輯特徵向量100: memory unit 110: The first circuit 120: second circuit 200: Memory device 600: Neural Network T1: the first switching element T2: second switching element R1: the first resistance element R2: second resistance element S301-S309: steps ML: match line SL1: search line BL: character line Ts: source Td: drain Tg: gate Tb: base Tf: floating gate Ti: insulating layer Te: Dielectric layer Tm: Metal layer L1: Curve Ln: Curve H1: Curve Hn: Curve LRS: low impedance state HRS: high impedance state ORS: initial impedance state TCAM1-TCAM8: Memory database CV1-CV4: Neural Analysis Unit Ly: connection layer LSH: Logical feature vector
第1A圖為根據本揭示內容之部份實施例之記憶單元的示意圖。 第1B圖為根據本揭示內容之部份實施例之記憶體裝置的示意圖。 第2A圖為根據本揭示內容之部份實施例之記憶單元的等效電路圖。 第2B圖為根據本揭示內容之部份實施例之開關元件的半導體結構示意圖。 第3圖為根據本揭示內容之部份實施例之記憶單元操作方法的步驟流程圖。 第4圖為根據本揭示內容之部份實施例之記憶單元之阻抗特性圖。 第5圖為根據本揭示內容之部份實施例之第一電路/第二電路的半導體結構示意圖。 第6圖為根據本揭示內容之部份實施例之記憶單元的應用示意圖。 FIG. 1A is a schematic diagram of a memory cell according to some embodiments of the present disclosure. FIG. 1B is a schematic diagram of a memory device according to some embodiments of the present disclosure. FIG. 2A is an equivalent circuit diagram of a memory cell according to some embodiments of the present disclosure. FIG. 2B is a schematic diagram of the semiconductor structure of the switching element according to some embodiments of the present disclosure. FIG. 3 is a flow chart of the steps of the memory unit operation method according to some embodiments of the present disclosure. FIG. 4 is an impedance characteristic diagram of a memory cell according to some embodiments of the present disclosure. FIG. 5 is a schematic diagram of the semiconductor structure of the first circuit/second circuit according to some embodiments of the present disclosure. FIG. 6 is a schematic diagram of the application of the memory unit according to some embodiments of the present disclosure.
國內寄存資訊(請依寄存機構、日期、號碼順序註記) 無 國外寄存資訊(請依寄存國家、機構、日期、號碼順序註記) 無 Domestic deposit information (please note in the order of deposit institution, date and number) no Foreign hosting information (please note in the order of hosting country, institution, date and number) no
S301-S309:步驟 S301-S309: steps
Claims (10)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109113475A TWI717269B (en) | 2020-04-22 | 2020-04-22 | Memory unit operation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109113475A TWI717269B (en) | 2020-04-22 | 2020-04-22 | Memory unit operation method |
Publications (2)
Publication Number | Publication Date |
---|---|
TWI717269B true TWI717269B (en) | 2021-01-21 |
TW202141488A TW202141488A (en) | 2021-11-01 |
Family
ID=75237223
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW109113475A TWI717269B (en) | 2020-04-22 | 2020-04-22 | Memory unit operation method |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI717269B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI800966B (en) * | 2021-10-29 | 2023-05-01 | 瑞昱半導體股份有限公司 | Content addressable memory cell |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5663905A (en) * | 1994-08-30 | 1997-09-02 | Mitsubishi Denki Kabushiki Kaisha | Semiconductor memory device comprising two kinds of memory cells operating in different access speeds and methods of operating and manufacturing the same |
US6724650B2 (en) * | 2001-10-22 | 2004-04-20 | Nec Electronics Corporation | Semiconductor device having a load less four transistor cell |
US7057460B2 (en) * | 2004-06-29 | 2006-06-06 | Rambus, Inc. | Differential amplifier with adaptive biasing and offset cancellation |
CN101064189A (en) * | 2006-04-24 | 2007-10-31 | 国际商业机器公司 | Static random access memory cell with improved stability and its formation method |
US8040718B2 (en) * | 2008-09-25 | 2011-10-18 | Kabushiki Kaisha Toshiba | Semiconductor memory device |
-
2020
- 2020-04-22 TW TW109113475A patent/TWI717269B/en not_active IP Right Cessation
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5663905A (en) * | 1994-08-30 | 1997-09-02 | Mitsubishi Denki Kabushiki Kaisha | Semiconductor memory device comprising two kinds of memory cells operating in different access speeds and methods of operating and manufacturing the same |
US6724650B2 (en) * | 2001-10-22 | 2004-04-20 | Nec Electronics Corporation | Semiconductor device having a load less four transistor cell |
US7057460B2 (en) * | 2004-06-29 | 2006-06-06 | Rambus, Inc. | Differential amplifier with adaptive biasing and offset cancellation |
CN101064189A (en) * | 2006-04-24 | 2007-10-31 | 国际商业机器公司 | Static random access memory cell with improved stability and its formation method |
US8040718B2 (en) * | 2008-09-25 | 2011-10-18 | Kabushiki Kaisha Toshiba | Semiconductor memory device |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI800966B (en) * | 2021-10-29 | 2023-05-01 | 瑞昱半導體股份有限公司 | Content addressable memory cell |
Also Published As
Publication number | Publication date |
---|---|
TW202141488A (en) | 2021-11-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
TWI699767B (en) | Circuitry for writing to and reading from an array of resistive random access memory cells | |
US10755780B2 (en) | Memory sense amplifier with precharge | |
US7978499B2 (en) | Semiconductor storage device | |
US7855923B2 (en) | Write current compensation using word line boosting circuitry | |
US9847132B1 (en) | Ternary content addressable memories | |
US20070159870A1 (en) | Nonvolatile semiconductor memory device | |
US8659926B1 (en) | PMC-based non-volatile CAM | |
EP3136398B1 (en) | Ternary content addressable memory (tcam) with programmable resistive elements | |
US10783976B2 (en) | Antifuse memory device and operation method thereof | |
CN107808680B (en) | Storage device | |
US20180040374A1 (en) | Ternary content addressable memories having a bit cell with memristors and serially connected match-line transistors | |
EP3785263A1 (en) | Method, system and device for integration of volatile and non-volatile memory bitcells | |
CN109584932B (en) | Memory device and operation method thereof | |
US9830996B2 (en) | Efuse bit cell, and read/write method thereof, and efuse array | |
US11837287B2 (en) | Memory sense amplifier with precharge | |
US10796741B1 (en) | Non-volatile memory with a select gate regulator circuit | |
JP4388889B2 (en) | MEMORY HAVING PRECHARGE CIRCUIT AND METHOD FOR PRECHARGING THE SAME | |
TWI717269B (en) | Memory unit operation method | |
US11495294B2 (en) | Hybrid self-tracking reference circuit for RRAM cells | |
US20220238154A1 (en) | Low-consumption rram memory differential reading | |
TWI797547B (en) | memory device | |
CN114496031A (en) | Memory device architecture using multiple cells per bit | |
TWI778928B (en) | Memory device and method of operating the same | |
US11984162B2 (en) | Hybrid self-tracking reference circuit for RRAM cells | |
US11915749B2 (en) | Resistive memory device and forming method thereof with improved forming time and improved forming uniformity |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
MM4A | Annulment or lapse of patent due to non-payment of fees |