TWI807542B - Memory system - Google Patents

Memory system Download PDF

Info

Publication number
TWI807542B
TWI807542B TW110147330A TW110147330A TWI807542B TW I807542 B TWI807542 B TW I807542B TW 110147330 A TW110147330 A TW 110147330A TW 110147330 A TW110147330 A TW 110147330A TW I807542 B TWI807542 B TW I807542B
Authority
TW
Taiwan
Prior art keywords
aforementioned
data
address
memory device
semiconductor memory
Prior art date
Application number
TW110147330A
Other languages
Chinese (zh)
Other versions
TW202326742A (en
Inventor
佐藤貴彦
Original Assignee
華邦電子股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 華邦電子股份有限公司 filed Critical 華邦電子股份有限公司
Priority to TW110147330A priority Critical patent/TWI807542B/en
Application granted granted Critical
Publication of TWI807542B publication Critical patent/TWI807542B/en
Publication of TW202326742A publication Critical patent/TW202326742A/en

Links

Landscapes

  • Techniques For Improving Reliability Of Storages (AREA)
  • Iron Core Of Rotating Electric Machines (AREA)
  • Soundproofing, Sound Blocking, And Sound Damping (AREA)
  • Vehicle Body Suspensions (AREA)

Abstract

Provide a memory system in which semiconductor memory device can be accessed properly. A memory system includes: a memory controller and a semiconductor memory device. The memory controller transmits an operation and an address, and a first checking data to the semiconductor memory device. When the semiconductor memory device receives a first response information that indicating no error detected, it transmits or receives read data or write data from the semiconductor memory device. When the semiconductor memory device receives the operation and the address, and the first checking data, it uses the first checking data to detect errors in the operation and the address, and transmits the first reply information when no error detected, and when no error detected in the operation or the address, transmits or receives read data or write data from the semiconductor memory device.

Description

記憶體系統memory system

本發明係有關包括記憶體控制器及半導體記憶裝置的記憶體系統。The present invention relates to a memory system including a memory controller and a semiconductor memory device.

已知傳統的記憶體系統包括:動態隨機存取記憶體(Dynamic Random Access Memory,DRAM)等之半導體記憶裝置,以及對該半導體記憶裝置進行讀取或寫入等控制的記憶體控制器(例如專利文獻1:特開2002-63791號公報)。半導體記憶裝置被配置為從記憶體控制器接收包含指令以及位址的指令包時,根據指令之內容(例如讀取指令、寫入指令)進行對應位址的記憶區域(例如記憶胞)之資料的讀取或寫入。Known conventional memory systems include: semiconductor memory devices such as Dynamic Random Access Memory (DRAM), and memory controllers for controlling reading or writing of the semiconductor memory devices (for example, Patent Document 1: JP-A-2002-63791). The semiconductor memory device is configured to read or write data in a memory area (such as a memory cell) corresponding to the address according to the content of the command (such as a read command or a write command) when receiving a command packet including a command and an address from a memory controller.

然而,傳統的記憶體系統中,從記憶體控制器往半導體記憶裝置的指令包傳送中時,若在指令或位址中發生錯誤,會有指令之內容或位址被變更的情況。因此,會有變得難以對半導體記憶裝置適當地存取之虞。However, in a conventional memory system, if an error occurs in a command or an address when the command packet is transmitted from the memory controller to the semiconductor memory device, the content of the command or the address may be changed. Therefore, it may become difficult to properly access the semiconductor memory device.

本發明提供之記憶體系統包括:記憶體控制器;以及半導體記憶裝置;前述記憶體控制器,被配置為進行:將指令及位址,以及前述指令及前述位址之錯誤偵測用之第一檢查資料傳送到前述半導體記憶裝置;以及從前述半導體記憶裝置接收顯示在前述指令及前述位址沒有偵測到錯誤之第一回應資訊時,與前述半導體記憶裝置之間根據前述指令傳送或接收對前述位址讀取或寫入之資料;前述半導體記憶裝置,被配置為進行:從前述記憶體控制器接收前述指令及前述位址,以及前述第一檢查資料時,利用前述第一檢查資料進行前述指令及前述位址之錯誤偵測,在前述指令及前述位址沒有偵測到錯誤時,將前述第一回應資訊傳送到前述記憶體控制器;以及在前述指令及前述位址沒有偵測到錯誤時,與前述記憶體控制器之間根據前述指令傳送或接收對前述位址讀取或寫入之資料。The memory system provided by the present invention includes: a memory controller; and a semiconductor memory device; the aforementioned memory controller is configured to: transmit instructions and addresses, and first inspection data for error detection of the aforementioned instructions and the aforementioned addresses to the aforementioned semiconductor memory device; When the command and the aforementioned address, and the aforementioned first check data, use the aforementioned first check data to perform error detection of the aforementioned command and the aforementioned address, when no error is detected in the aforementioned command and the aforementioned address, the aforementioned first response information is sent to the aforementioned memory controller;

根據本發明之記憶體系統,可以對半導體記憶裝置適當地存取。According to the memory system of the present invention, it is possible to properly access the semiconductor memory device.

(第一實施型態) 如第1圖所示,記憶體系統包括記憶體控制器10以及半導體記憶裝置20。在本實施型態中,以半導體記憶裝置20為虛擬靜態隨機存取記憶體(pseudo-Static Random Access Memory,pSRAM)作為一例進行說明。此處,pSRAM是包括與靜態隨機存取記憶體(Static Random Access Memory,SRAM)有相容性的介面的半導體記憶裝置。另外,pSRAM是將動態隨機存取記憶體(Dynamic Random Access Memory,DRAM)作為記憶胞記憶資料,重新設計DRAM之存取介面,使之與SRAM之存取介面有相容性。在本實施型態中,以半導體記憶裝置20為和時脈訊號同步接收訊號的時脈同步型pSRAM,亦為位址資料多工介面型pSRAM的情況作為一例展示。位址資料多工介面型pSRAM具有被配置為輸入每個位址訊號以及資料訊號的位址資料端子。另外,在本實施型態的pSRAM中,雖然以採用雙倍資料速率(Double Data Rate,DDR)作為資料傳送方式的情況作為一例展示,pSRAM也可以採用單一資料速率(Single Data Rate,SDR)。 (first implementation type) As shown in FIG. 1 , the memory system includes a memory controller 10 and a semiconductor memory device 20 . In this embodiment, it is described that the semiconductor memory device 20 is a pseudo-Static Random Access Memory (pseudo-Static Random Access Memory, pSRAM) as an example. Here, pSRAM is a semiconductor memory device including an interface compatible with Static Random Access Memory (SRAM). In addition, pSRAM uses Dynamic Random Access Memory (DRAM) as memory cell memory data, and redesigns the access interface of DRAM to make it compatible with the access interface of SRAM. In this embodiment, a case where the semiconductor memory device 20 is a clock synchronous pSRAM that receives signals synchronously with a clock signal and is also an address data multiplex interface type pSRAM is shown as an example. The address-data multiplexer type pSRAM has address-data terminals configured to input each address signal and data signal. In addition, in the pSRAM of this embodiment, although the case of adopting double data rate (Double Data Rate, DDR) as the data transmission method is shown as an example, the pSRAM can also adopt single data rate (Single Data Rate, SDR).

記憶體控制器10被配置為進行:將指令(在圖中顯示為Operation)及位址(在圖中顯示為Address),以及指令及位址的錯誤偵測用之第一檢查資料(在圖中顯示為CheckA)傳送到半導體記憶裝置20;以及從前述半導體記憶裝置20接收顯示在指令及位址中沒有偵測到錯誤之第一回應資訊(在圖中顯示為ReplyA(OK))時,與半導體記憶裝置20之間根據指令傳送或接收對位址讀取或寫入之資料(在圖中顯示為Data)。另一方面,半導體記憶裝置20被配置為進行:從記憶體控制器10接收指令及位址,以及第一檢查資料時,利用第一檢查資料進行指令及位址之錯誤偵測,在指令及位址沒有偵測到錯誤時,將第一回應資訊(ReplyA(OK))傳送到記憶體控制器10,與記憶體控制器之間10根據指令傳送或接收對位址讀取或寫入之資料 。The memory controller 10 is configured to: transmit an instruction (shown as Operation in the figure) and an address (shown as Address in the figure), and first check data (shown as CheckA in the figure) for error detection of the command and address to the semiconductor memory device 20; Or written data (shown as Data in the figure). On the other hand, the semiconductor memory device 20 is configured to perform: when receiving instructions and addresses from the memory controller 10, and the first check data, use the first check data to detect errors in the instructions and addresses, and when no errors are detected in the instructions and addresses, send the first response information (ReplyA (OK)) to the memory controller 10, and transmit or receive data read or written to the address according to the instructions between the memory controller 10.

另外,半導體記憶裝置20被配置為:在指令及位址之任一者偵測到錯誤時,將顯示在指令及位址之任一者偵測到錯誤的第二回應資訊(在圖中顯示為ReplyA(NG))傳送到記憶體控制器10。另一方面,記憶體控制器10被配置為:從半導體記憶裝置20接收第二回應資訊(ReplyA(NG))時,將指令及位址再傳送到半導體記憶裝置20,可以對半導體記憶裝置20從最初開始重新存取。In addition, the semiconductor memory device 20 is configured to: when any one of the command and the address detects an error, transmit the second response information (shown as ReplyA(NG) in the figure) to the memory controller 10 indicating that any one of the command and the address has detected an error. On the other hand, the memory controller 10 is configured to: when receiving the second response information (ReplyA(NG)) from the semiconductor memory device 20, retransmit the command and the address to the semiconductor memory device 20, so that the semiconductor memory device 20 can be re-accessed from the beginning.

再者,記憶體控制器10(記憶體控制器10及半導體記憶裝置20中之一者) 被配置為:將寫入之資料以及該資料之錯誤偵測用之第二檢查資料(在圖中顯示為CheckD),傳送到半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之另一者)。另一方面,半導體記憶裝置20被配置為:接收資料以及第二檢查資料(CheckD)時,利用第二檢查資料進行資料之錯誤偵測,在資料中沒有偵測到錯誤時,將顯示沒有在資料中偵測到錯誤的第三回應資訊(在圖中顯示為ReplyD(OK))傳送到記憶體控制器10。在此情況下,由於可以抑制維持在讀取或寫入之資料中包含錯誤的狀態,可以確保資料的完整性。Moreover, the memory controller 10 (one of the memory controller 10 and the semiconductor memory device 20) is configured to: transmit the written data and the second check data (shown as CheckD in the figure) for error detection of the data to the semiconductor memory device 20 (the other of the memory controller 10 and the semiconductor memory device 20). On the other hand, the semiconductor memory device 20 is configured to: when receiving the data and the second check data (CheckD), use the second check data to perform data error detection, and when no error is detected in the data, send the third response information (shown as ReplyD (OK) in the figure) to the memory controller 10 indicating that no error has been detected in the data. In this case, since it is suppressed to maintain a state in which an error is included in the read or written data, the integrity of the data can be ensured.

再者還有,半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之另一者)被配置為進行:在資料中偵測到錯誤時,將顯示在資料中偵測到錯誤的第四回應資訊(在圖中顯示為ReplyD(NG))傳送到記憶體控制器10(記憶體控制器10及半導體記憶裝置20中之一者);以及直到經過特定時間為止,等待從記憶體控制器10再傳送資料,可以繼續進行有關該資料之寫入。因此,和從最初開始重新存取有關該資料之寫入的情況(再傳送指令及位址的情況) 比較之下,可以提高存取效率。另一方面,記憶體控制器10(記憶體控制器10及半導體記憶裝置20中之一者)被配置為:接收第四回應資訊(ReplyD(NG))時,將該資料以及對應該資料之第二檢查資料再傳送到半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之另一者)。因此,在資料中偵測到錯誤時,藉由記憶體控制器10再傳送資料,可以再度進行該資料之寫入。Furthermore, the semiconductor memory device 20 (the other of the memory controller 10 and the semiconductor memory device 20) is configured to perform: when an error is detected in the data, the fourth response information (shown as ReplyD (NG) shown in the figure) is transmitted to the memory controller 10 (one of the memory controller 10 and the semiconductor memory device 20 ) indicating that an error has been detected in the data; Therefore, compared with the case of re-accessing the writing of the data from the beginning (the case of re-transferring the command and address), the access efficiency can be improved. On the other hand, the memory controller 10 (one of the memory controller 10 and the semiconductor memory device 20) is configured to: when receiving the fourth response information (ReplyD(NG)), send the data and the second inspection data corresponding to the data to the semiconductor memory device 20 (the other of the memory controller 10 and the semiconductor memory device 20). Therefore, when an error is detected in the data, the memory controller 10 retransmits the data, so that the data can be written again.

再者,半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之另一者)被配置為:直到經過特定時間為止,沒有從記憶體控制器10(記憶體控制器10及半導體記憶裝置20中之一者)再傳送資料時,解除等待資料之再傳送。因此,在沒有再傳送資料的情況下,藉由解除等待資料之再傳送,半導體記憶裝置20可以處理別的讀取或寫入存取,可以提高存取的處理效率。Furthermore, the semiconductor memory device 20 (the other one of the memory controller 10 and the semiconductor memory device 20) is configured to cancel waiting for retransmission of data when data is not retransmitted from the memory controller 10 (one of the memory controller 10 and the semiconductor memory device 20) until a specific time elapses. Therefore, in the case of no retransmission of data, the semiconductor memory device 20 can process other read or write accesses by canceling the retransmission of waiting data, and the processing efficiency of access can be improved.

記憶體控制器10連接到外部的主機裝置(圖示中省略),回應來自主機裝置的命令,對半導體記憶裝置20進行讀取或寫入等之存取。例如,記憶體控制器10從主機裝置接收用以對半導體記憶裝置20進行資料之寫入的寫入要求,或用以從半導體記憶裝置20進行資料之讀取的讀取要求時,使晶片選擇訊號/CE有效(低位準)並傳送到半導體記憶裝置20。另外,記憶體控制器10將從主機裝置接收之時脈訊號作為外部時脈訊號CLK,傳送到半導體記憶裝置20。The memory controller 10 is connected to an external host device (not shown in the figure), responds to commands from the host device, and accesses the semiconductor memory device 20 such as reading or writing. For example, when the memory controller 10 receives a write request for writing data into the semiconductor memory device 20 from the host device, or a read request for reading data from the semiconductor memory device 20, the chip select signal /CE is enabled (low level) and sent to the semiconductor memory device 20. In addition, the memory controller 10 transmits the clock signal received from the host device to the semiconductor memory device 20 as the external clock signal CLK.

記憶體控制器10包括:要求控制部11以及序列控制部12。要求控制部11及序列控制部12也可以由專用之硬體裝置或邏輯電路等構成。The memory controller 10 includes a request control unit 11 and a sequence control unit 12 . The request control unit 11 and the sequence control unit 12 may also be constituted by dedicated hardware devices or logic circuits.

記憶體控制器10也可以包括:中央處理器(Central Processing Unit,CPU)(圖示中省略)、例如隨機存取記憶體(Random Access Memory,RAM)等記憶裝置(圖示中省略)、半導體記憶裝置20以及用以在各主機裝置之間進行訊號之傳送接收的介面(圖示中省略)。舉例而言,記憶體控制器10之CPU也可以藉由讀取並執行記憶於記憶裝置的電腦程式,實現要求控制部11及序列控制部12之機能。The memory controller 10 may also include: a central processing unit (Central Processing Unit, CPU) (omitted in the figure), memory devices such as random access memory (Random Access Memory, RAM) (omitted in the figure), a semiconductor memory device 20, and an interface (omitted in the figure) for transmitting and receiving signals between host devices. For example, the CPU of the memory controller 10 can also realize the functions of the request control unit 11 and the sequence control unit 12 by reading and executing the computer program stored in the memory device.

要求控制部11被配置為從主機裝置接收對半導體記憶裝置20寫入或讀取之要求時,生成用以變換記憶體控制器10與半導體記憶裝置20之間傳送或接收之位址資料訊號ADQ的傳送方式的變換訊號。具體而言,要求控制部11從主機裝置接收用以對半導體記憶裝置20進行資料之寫入的寫入要求時,將變換訊號deson輸出到序列控制部12,該變換訊號deson用以將從記憶體控制器10傳送到半導體記憶裝置20的位址資料訊號ADQ,從串列傳輸方式變換(反串列化)為特定數之位元(例如8位元)的並列傳輸方式。The request control unit 11 is configured to generate a conversion signal for converting the transmission mode of the address data signal ADQ transmitted or received between the memory controller 10 and the semiconductor memory device 20 when receiving a request for writing or reading from the semiconductor memory device 20 from the host device. Specifically, when the request control unit 11 receives a write request for writing data into the semiconductor memory device 20 from the host device, it outputs the conversion signal deson to the sequence control unit 12. The conversion signal deson is used to convert (de-serialize) the address data signal ADQ transmitted from the memory controller 10 to the semiconductor memory device 20 from the serial transmission method to the parallel transmission method of a specific number of bits (for example, 8 bits).

另外,要求控制部11從主機裝置接收用以對半導體記憶裝置20進行資料之讀取的讀取要求時,將變換訊號seron輸出到序列控制部12,該變換訊號seron用以將從半導體記憶裝置20傳送到記憶體控制器10之訊號(例如讀取資料等)從並列傳輸方式變換(串列化)為串列傳輸方式。In addition, when the request control unit 11 receives a read request for reading data from the semiconductor memory device 20 from the host device, it outputs a conversion signal seron to the sequence control unit 12. The conversion signal seron is used to convert (serialize) the signal (such as reading data) transmitted from the semiconductor memory device 20 to the memory controller 10 from the parallel transmission method to the serial transmission method.

序列控制部12被配置為基於變換訊號deson變換位址資料訊號ADQ的傳送方式。另外,序列控制部12包括:串聯器/解串器(Serializer/deserializer)(SerDes)12a以及錯誤控制部12b。SerDes12a使在記憶體控制器10及半導體記憶裝置20之間傳送或接收之訊號,在串列傳輸方式及並列傳輸方式之間互相變換。另外,雖然在本實施型態中,以SerDes 12a反串列化寫入資料 ,並串列化讀取資料的情況做為一例進行說明,SerDes 12a也可以被配置為串列化寫入資料,例如從64位元到8位元等,並反串列化讀取資料。另外,在本實施型態中,序列控制部12為本發明之「第一序列控制部」的一例,SerDes 12a為本發明之「第一串聯器/解串器」的一例。The sequence control unit 12 is configured to convert the transmission mode of the address data signal ADQ based on the conversion signal deson. In addition, the sequence control unit 12 includes a serializer/deserializer (SerDes) 12a and an error control unit 12b. The SerDes 12a converts the signals transmitted or received between the memory controller 10 and the semiconductor memory device 20 between the serial transmission method and the parallel transmission method. In addition, although in this embodiment, the case where the SerDes 12a deserializes the written data and serializes the read data is taken as an example for illustration, the SerDes 12a can also be configured to serialize the written data, such as from 64 bits to 8 bits, etc., and deserializes the read data. In addition, in this embodiment, the serial control unit 12 is an example of the "first serial control unit" of the present invention, and the SerDes 12a is an example of the "first serializer/deserializer" of the present invention.

舉例而言,在要求控制部11接收寫入要求或讀取要求的情況下,在變換訊號deson從要求控制部11輸入時,SerDes 12a生成讀取指令或寫入指令(Operation)以及位址(Address),並輸出到錯誤控制部12b。接著,SerDes 12a在指令及位址(Address)之錯誤偵測用之第一檢查資料(CheckA)從錯誤控制部12b輸入時,將指令、位址及第一檢查資料變換為並列傳輸方式。接著,SerDes 12a將指令、位址及第一檢查資料作為位址資料訊號ADQ傳送到半導體記憶裝置20。For example, when the request control unit 11 receives a write request or a read request, when the converted signal deson is input from the request control unit 11, the SerDes 12a generates a read command or a write command (Operation) and an address (Address), and outputs them to the error control unit 12b. Next, SerDes 12a converts the command, address, and first check data into parallel transmission when the first check data (CheckA) for error detection of command and address (Address) is input from error control unit 12b. Next, the SerDes 12a transmits the command, address and first check data to the semiconductor memory device 20 as an address data signal ADQ.

另外,SerDes 12a從半導體記憶裝置20接收作為位址資料訊號ADQ之顯示在寫入指令及位址沒有偵測到錯誤的第一回應資訊(ReplyA(OK))的情況下,將從主機裝置接收之寫入資料(Data)輸出到錯誤控制部12b。接著,SerDes 12a在寫入資料之錯誤偵測用之第二檢查資料從錯誤控制部12b輸入時,第二檢查資料及寫入資料變換為並列傳輸方式。接著,SerDes 12a將第二檢查資料及寫入資料作為位址資料訊號ADQ傳送到半導體記憶裝置20。In addition, the SerDes 12a receives from the semiconductor memory device 20 as the address data signal ADQ the first response information (ReplyA (OK)) indicating that there is no error detected in the write command and the address, and outputs the write data (Data) received from the host device to the error control unit 12b. Next, when the SerDes 12a receives the second inspection data for error detection of the written data from the error control unit 12b, the second inspection data and the written data are converted to a parallel transmission method. Then, the SerDes 12a sends the second check data and write data to the semiconductor memory device 20 as the address data signal ADQ.

再者,SerDes 12a從半導體記憶裝置20接收作為位址資料訊號ADQ之顯示在寫入指令及位址中之任一者偵測到錯誤的第二回應資訊(ReplyA(NG)) 的情況下,在用以再傳送寫入指令、位址及第一檢查資料之訊號retrywr從錯誤控制部12b輸入時,將寫入指令 、位址及第一檢查資料作為位址資料訊號ADQ再傳送到半導體記憶裝置20。Moreover, SerDes 12a receives from the semiconductor memory device 20 as the address data signal ADQ the second response information (ReplyA(NG)) indicating that any one of the write command and the address detects an error.

再者還有,SerDes 12a從半導體記憶裝置20接收作為位址資料訊號ADQ之顯示在讀取指令及位址中之任一者偵測到錯誤的第二回應資訊(ReplyA(NG)) 的情況下,在用以再傳送讀取指令、位址及第一檢查資料之訊號retryrd從錯誤控制部12b輸入時,將讀取指令 、位址及第一檢查資料作為位址資料訊號ADQ再傳送到半導體記憶裝置20。Furthermore, SerDes 12a receives from the semiconductor memory device 20 as the address data signal ADQ the second response information (ReplyA(NG)) indicating that any one of the read command and the address detects an error.

另外,SerDes 12a從半導體記憶裝置20接收作為位址資料訊號ADQ之顯示在寫入資料偵測到錯誤的第四回應資訊(ReplyD(NG)) 的情況下,將寫入資料及對應該寫入資料之第二檢查資料作為位址資料訊號ADQ再傳送到半導體記憶裝置20。In addition, the SerDes 12a receives from the semiconductor memory device 20 as the address data signal ADQ the fourth response information (ReplyD(NG)) indicating that an error is detected in the write data, and then sends the write data and the second check data corresponding to the write data to the semiconductor memory device 20 as the address data signal ADQ.

舉例而言,第二檢查資料與第一檢查資料可以被配置為同位碼或循環冗餘校驗 (Cyclic Redundancy Checking,CRC)碼等。For example, the second check data and the first check data can be configured as parity codes or cyclic redundancy check (Cyclic Redundancy Checking, CRC) codes and the like.

接著說明半導體記憶裝置20包括:序列控制部21;指令控制部22;位址控制部23;字元線控制部24;行控制部25;資料控制部26;資料匯流排控制部27;感應放大器28以及記憶胞陣列29。半導體記憶裝置20內之各部21~29也可以由專用之硬體裝置或邏輯電路構成。Next, the semiconductor memory device 20 includes: a sequence control unit 21; a command control unit 22; an address control unit 23; a word line control unit 24; a line control unit 25; a data control unit 26; a data bus control unit 27; The various parts 21-29 in the semiconductor memory device 20 may also be constituted by dedicated hardware devices or logic circuits.

序列控制部21被配置為變換與記憶體控制器10之間傳送接收之位址資料訊號ADQ的傳送方式。另外,序列控制部21包括:SerDes 21a以及錯誤控制部21b。SerDes 21a與記憶體控制器10之SerDes 12a同樣地,使在記憶體控制器10及半導體記憶裝置20之間傳送或接收之訊號,在串列傳輸方式及並列傳輸方式之間互相變換。另外,序列控制部21被配置為將資料選通訊號RWDS傳送到記憶體控制器10之序列控制部12。另外,在本實施型態中,序列控制部21為本發明之「第二序列控制部」之一例,SerDes 21a為本發明之「第二串聯器/解串器」的一例,錯誤控制部21b為本發明之「第二錯誤控制部」之一例。The sequence control unit 21 is configured to convert the transmission mode of the address data signal ADQ transmitted and received between the memory controller 10 . In addition, the sequence control unit 21 includes a SerDes 21a and an error control unit 21b. Like the SerDes 12a of the memory controller 10, the SerDes 21a converts the signals transmitted or received between the memory controller 10 and the semiconductor memory device 20 between the serial transmission method and the parallel transmission method. In addition, the sequence control unit 21 is configured to transmit the data strobe signal RWDS to the sequence control unit 12 of the memory controller 10 . In addition, in this embodiment, the sequence control unit 21 is an example of the “second sequence control unit” of the present invention, the SerDes 21a is an example of the “second serializer/deserializer” of the present invention, and the error control unit 21b is an example of the “second error control unit” of the present invention.

SerDes 21a從記憶體控制器10接收作為位址資料訊號ADQ之指令、位址及第一檢查資料時,將指令、位址及第一檢查資料變換為串列傳輸方式輸出到錯誤控制部21b。錯誤控制部21b利用第一檢查資料進行指令及位址之錯誤偵測。在沒有從指令及位址(Adress)偵測到錯誤時,錯誤控制部21b將顯示在指令及位址(Adress)中沒有偵測到錯誤之第一回應資訊(ReplyA(OK))輸出到SerDes 21a。When the SerDes 21a receives the command, address and first check data as the address data signal ADQ from the memory controller 10, it converts the command, address and first check data into a serial transmission mode and outputs it to the error control unit 21b. The error control unit 21b uses the first check data to detect errors of commands and addresses. When no error is detected from the command and address (Adress), the error control unit 21b outputs the first response information (ReplyA (OK)) indicating that no error is detected in the command and address (Adress) to the SerDes 21a.

SerDes 21a在顯示在指令及位址中沒有偵測到錯誤之第一回應資訊(ReplyA(OK))從錯誤控制部21b輸入時,將第一回應資訊(ReplyA(OK))變換為並行傳輸方式,將第一回應資訊(ReplyA(OK))作為位址資料訊號ADQ傳送到記憶體控制器10。另外,在此情況下,SerDes 21a將顯示接收之指令之內容(讀取指令或寫入指令)的訊號ope輸出到指令控制部22,將顯示接收之位址的訊號adr輸出到位址控制部23。SerDes 21a converts the first response information (ReplyA(OK)) into a parallel transmission method when the first response information (ReplyA(OK)) indicating that no error is detected in the command and address is input from the error control unit 21b, and transmits the first response information (ReplyA(OK)) to the memory controller 10 as the address data signal ADQ. In addition, in this case, the SerDes 21a outputs a signal ope indicating the content of the received command (read command or write command) to the command control unit 22 and a signal adr indicating the received address to the address control unit 23 .

另一方面,錯誤控制部21b在指令及位址(Adress)中之任一者偵測到錯誤時,將顯示在指令及位址(Adress)中之任一者偵測到錯誤的第二回應資訊(ReplyA(NG)) 輸出到SerDes 21a。SerDes 21a在顯示在指令及位址中偵測到錯誤之第二回應資訊(ReplyA(NG))從錯誤控制部21b輸入時,將第二回應資訊(ReplyA(NG))變換為並行傳輸方式,將第二回應資訊(ReplyA(NG))作為位址資料訊號ADQ傳送到記憶體控制器10。On the other hand, when an error is detected in any one of the command and the address (Adress), the error control unit 21b outputs the second response information (ReplyA(NG)) indicating that any one of the command and the address (Adress) has detected an error to the SerDes 21a. SerDes 21a converts the second response information (ReplyA(NG)) into a parallel transmission method when the second response information (ReplyA(NG)) indicating an error detected in the command and address is input from the error control unit 21b, and transmits the second response information (ReplyA(NG)) to the memory controller 10 as the address data signal ADQ.

再者還有,SerDes 21a在從記憶體控制器10接收作為位址資料訊號ADQ之第二檢查資料及寫入資料時,將第二檢查資料及寫入資料變換為串列傳輸方式輸出到錯誤控制部21b。錯誤控制部21b利用第二檢查資料進行寫入資料之錯誤偵測。沒有從寫入資料偵測到錯誤時,錯誤控制部21b將顯示在寫入資料中沒有偵測到錯誤之第三回應資訊(ReplyD(OK))輸出到SerDes 21a。Furthermore, when the SerDes 21a receives the second check data and write data as the address data signal ADQ from the memory controller 10, it converts the second check data and write data into serial transmission and outputs them to the error control unit 21b. The error control unit 21b uses the second check data to detect errors in writing data. When no error is detected from the written data, the error control unit 21b outputs third response information (ReplyD(OK)) indicating that no error has been detected in the written data to the SerDes 21a.

SerDes 21a在顯示在寫入資料中沒有偵測到錯誤之第三回應資訊(ReplyD(OK))從錯誤控制部21b輸入時,將第三回應資訊(ReplyD(OK))變換為並行傳輸方式,將第三回應資訊(ReplyD(OK))作為位址資料訊號ADQ傳送到記憶體控制器10。另外,在此情況下,SerDes 21a將顯示接收之資料的訊號di輸出到資料控制部26。SerDes 21a converts the third response information (ReplyD(OK)) into a parallel transmission method when the third response information (ReplyD(OK)) indicating that no error is detected in the written data is input from the error control unit 21b, and transmits the third response information (ReplyD(OK)) to the memory controller 10 as the address data signal ADQ. In addition, in this case, the SerDes 21a outputs a signal di indicating received data to the data control unit 26 .

另一方面,錯誤控制部21b從寫入資料偵測到錯誤時,將顯示在寫入資料中偵測到錯誤之第四回應資訊(ReplyD(NG))輸出到SerDes 21a。SerDes 21a在顯示在寫入資料中偵測到錯誤之第四回應資訊(ReplyD(NG))從錯誤控制部21b輸入時,將第四回應資訊(ReplyD(NG))變換為並行傳輸方式,將第四回應資訊(ReplyD(NG))作為位址資料訊號ADQ傳送到記憶體控制器10。On the other hand, when an error is detected from the written data, the error control unit 21b outputs fourth response information (ReplyD(NG)) indicating that an error has been detected in the written data to the SerDes 21a. SerDes 21a converts the fourth response information (ReplyD(NG)) into a parallel transmission method when the fourth response information (ReplyD(NG)) indicating that an error is detected in the written data is input from the error control unit 21b, and transmits the fourth response information (ReplyD(NG)) to the memory controller 10 as an address data signal ADQ.

再者還有,SerDes 21a在顯示讀取資料之訊號do從資料控制部26輸入時,將讀取資料輸出到錯誤控制部21b。接著,SerDes 21a在從錯誤控制部21b輸入讀取資料之錯誤偵測用之第二檢查資料時,將讀取資料及第二檢查資料變換為並行傳輸方式,將讀取資料及第二檢查資料作為位址資料訊號ADQ傳送到記憶體控制器10。Furthermore, the SerDes 21a outputs the read data to the error control unit 21b when the signal do indicating the read data is input from the data control unit 26 . Next, when the SerDes 21a inputs the second check data for error detection of the read data from the error control unit 21b, it converts the read data and the second check data into a parallel transmission method, and transmits the read data and the second check data to the memory controller 10 as an address data signal ADQ.

再者還有,錯誤控制部21b在晶片選擇訊號/CE從無效(高位準)變化為有效(低位準)時,使顯示正在讀取或寫入存取之處理中的訊號busy有效(高位準)並輸出到指令控制部22。指令控制部22被配置為基於從記憶體控制器10輸入之指令生成內部指令。具體而言,指令控制部22在有效之訊號busy從序列控制部21輸入時,回應從序列控制部21輸入之訊號ope生成內部指令。此處,生成之內部指令舉例而言,包含讀取訊號、寫入訊號、更新訊號等。指令控制部22回應生成之內部指令,使用以活性化字元線之訊號wlon有效並輸出到字元線控制部24,使用以活性化位元線之訊號clon有效並輸出到行控制部25,使用以活性化感應放大器28之訊號saon有效並輸出到感應放大器28。另外,指令控制部22使用以非活性化字元線之訊號wloff有效並輸出到字元線控制部24。Furthermore, when the chip selection signal /CE changes from invalid (high level) to valid (low level), the error control unit 21b makes the signal busy (high level) indicating that the read or write access process is in progress and outputs it to the command control unit 22. The command control section 22 is configured to generate internal commands based on commands input from the memory controller 10 . Specifically, the command control unit 22 generates an internal command in response to the signal ope input from the sequence control unit 21 when a valid signal busy is input from the sequence control unit 21 . Here, the generated internal command includes, for example, a read signal, a write signal, and an update signal. The command control section 22 responds to the generated internal command, the signal wlon used to activate the word line is valid and output to the word line control section 24, the signal clon used to activate the bit line is valid and output to the row control section 25, and the signal saon used to activate the sense amplifier 28 is valid and output to the sense amplifier 28. In addition, the command control unit 22 uses the signal wloff for deactivating the word line to be valid and output to the word line control unit 24 .

位址控制部23被配置為基於從記憶體控制器10輸入之位址控制對應該位址的字元線以及位元線活性化。具體而言,位址控制部23在訊號adr從序列控制部21輸入時,生成顯示活性化之字元線的列位址訊號ra,以及顯示活性化之位元線的行位址訊號ca。接著,位址控制部23將生成之列位址訊號ra輸出到字元線控制部24,將生成之行位址訊號ca輸出到行控制部25。The address control unit 23 is configured to control activation of word lines and bit lines corresponding to the address based on the address input from the memory controller 10 . Specifically, the address control unit 23 generates a column address signal ra indicating an activated word line and a row address signal ca indicating an activated bit line when the signal adr is input from the sequence control unit 21 . Next, the address control unit 23 outputs the generated column address signal ra to the word line control unit 24 , and outputs the generated row address signal ca to the row control unit 25 .

字元線控制部24在訊號wlon為有效狀態下從指令控制部22輸入時,將由從位址控制部23輸入之列位址訊號ra顯示之用以活性化字元線之訊號wl輸出到記憶胞陣列29,活性化(驅動)該字元線。另外,字元線控制部24在訊號wloff為有效狀態下從指令控制部22輸入時,將用以非活性化字元線之訊號wl輸出到記憶胞陣列29,非活性化該字元線。When the word line control unit 24 receives input from the command control unit 22 when the signal wlon is active, it outputs the signal wl for activating the word line indicated by the column address signal ra input from the address control unit 23 to the memory cell array 29 to activate (drive) the word line. In addition, when the signal wloff is input from the command control unit 22 while the signal wloff is active, the word line control unit 24 outputs the signal wl for deactivating the word line to the memory cell array 29 to deactivate the word line.

行控制部25在訊號clon為有效狀態下從指令控制部22輸入時,在記憶胞陣列29內之複數個位元線之中,將由從位址控制部23輸入之行位址訊號ca顯示之用以活性化位元線之訊號cl輸出到感應放大器28。When the row control unit 25 receives input from the command control unit 22 when the signal clon is active, among the plurality of bit lines in the memory cell array 29, the signal cl for activating the bit lines indicated by the row address signal ca input from the address control unit 23 is output to the sense amplifier 28.

資料控制部26在顯示寫入資料之訊號di從序列控制部21輸入時,將顯示寫入資料之訊號wdb輸出到資料匯流排控制部27。另外,資料控制部26在顯示讀取資料之訊號rdb從資料匯流排控制部27輸入時,將顯示讀取資料之訊號do輸出到序列控制部21。The data control unit 26 outputs a signal wdb indicating write data to the data bus control unit 27 when the signal di indicating write data is input from the sequence control unit 21 . In addition, the data control unit 26 outputs the signal do indicating the read data to the sequence control unit 21 when the signal rdb indicating the read data is input from the data bus control unit 27 .

資料匯流排控制部27在顯示寫入資料之訊號wdb從資料控制部26輸入時,將顯示寫入資料之訊號cdb輸出到感應放大器28。另外,資料匯流排控制部27在顯示讀取資料之訊號cdb從感應放大器28輸入時,將顯示讀取資料之訊號rdb輸出到資料控制部26。The data bus control unit 27 outputs the signal cdb indicating the written data to the sense amplifier 28 when the signal wdb indicating the written data is input from the data control unit 26 . In addition, the data bus control unit 27 outputs the signal rdb indicating the read data to the data control unit 26 when the signal cdb indicating the read data is input from the sense amplifier 28 .

感應放大器28在訊號saon為有效狀態下從指令控制部22輸入時,將由從行控制部25輸入之訊號cl顯示之用以活性化位元線之訊號bl輸出到記憶胞陣列29,活性化(驅動)該位元線。接著,感應放大器28經由被活性化之位元線,對記憶胞進行資料之讀寫。舉例而言,感應放大器28將從資料匯流排控制部27輸入之顯示寫入資料之訊號cdb,經由被活性化之位元線寫入記憶胞。另外,感應放大器28經由被活性化之位元線,將從記憶胞讀取之資料輸出到資料匯流排控制部27。The sense amplifier 28 outputs the signal bl for activating the bit line shown by the signal cl input from the row control unit 25 to the memory cell array 29 when the signal saon is input from the command control unit 22 to activate (drive) the bit line. Next, the sense amplifier 28 reads and writes data to the memory cell through the activated bit line. For example, the sense amplifier 28 writes the signal cdb input from the data bus control part 27 indicating the write data into the memory cell through the activated bit line. In addition, the sense amplifier 28 outputs the data read from the memory cell to the data bus control unit 27 via the activated bit line.

記憶胞陣列29包含行列(Array)狀配置之複數個記憶胞(圖示中省略)。在各記憶胞中,記憶從記憶體控制器10輸入之資料。各記憶胞可以是習知的1T1C(1電晶體1電容)型記憶胞。另外,各記憶胞連接複數個字元線之中任一者之字元線,以及複數個位元線之中任一者之位元線。The memory cell array 29 includes a plurality of memory cells (not shown) arranged in an array. In each memory cell, data input from the memory controller 10 is stored. Each memory cell can be a conventional 1T1C (1 transistor 1 capacitor) type memory cell. In addition, each memory cell is connected to a word line of any one of the plurality of word lines and a bit line of any one of the plurality of bit lines.

參照第2圖,以寫入指令從記憶體控制器10輸入的情況作為一例進行說明。Referring to FIG. 2, a case where a write command is input from the memory controller 10 will be described as an example.

首先,半導體記憶裝置20從記憶體控制器10接收指令、位址及第一檢查資料時,利用第一檢查資料進行指令及位址之錯誤偵測。接著,半導體記憶裝置20在沒有偵測到錯誤時,將顯示在指令及位址中沒有偵測到錯誤之第一回應資訊(ReplyA(OK))傳送到記憶體控制器10。另外,第一回應資訊(ReplyA(OK))也可以在從指令及位址被輸入到資料被輸入或輸出為止的期間傳送。Firstly, when the semiconductor memory device 20 receives the command, the address and the first check data from the memory controller 10, it uses the first check data to detect the error of the command and the address. Then, when no error is detected, the semiconductor memory device 20 sends the first response information (ReplyA(OK)) indicating that no error is detected in the command and address to the memory controller 10 . In addition, the first response information (ReplyA(OK)) may be transmitted during the period from when a command and an address are input to when data is input or output.

另外,半導體記憶裝置20在指令及位址中沒有偵測到錯誤時,使訊號wlon有效(高位準)。在此情況下,使用以活性化對應位址之字元線的訊號wl有效(高位準)。In addition, the semiconductor memory device 20 makes the signal wlon active (high level) when no error is detected in the command and address. In this case, the signal wl is asserted (high level) used to activate the word line corresponding to the address.

另一方面,記憶體控制器10接收第一回應資訊(ReplyA(OK))時,將第二檢查資料及寫入資料傳送到半導體記憶裝置20。另外,第二檢查資料也可以在延遲期間傳送。On the other hand, when the memory controller 10 receives the first response information (ReplyA(OK)), it transmits the second checking data and writing data to the semiconductor memory device 20 . In addition, the second inspection data can also be transmitted during the delay.

半導體記憶裝置20在從記憶體控制器10接收第二檢查資料及寫入資料時,利用第二檢查資料進行寫入資料之錯誤偵測。另外,半導體記憶裝置20使對應位址之位元線的訊號cl有效(高位準)。此時,半導體記憶裝置20之感應放大器28藉由活性化對應位址之位元線,可以將寫入資料寫入記憶胞陣列29內之記憶胞。When the semiconductor memory device 20 receives the second check data and the write data from the memory controller 10 , it uses the second check data to detect errors in the write data. In addition, the semiconductor memory device 20 makes the signal cl of the bit line corresponding to the address valid (high level). At this time, the sense amplifier 28 of the semiconductor memory device 20 can write the writing data into the memory cells in the memory cell array 29 by activating the bit line corresponding to the address.

再者,半導體記憶裝置20在接收寫入資料後始晶片選擇訊號/CE無效(高位準)時,將顯示在寫入資料中是否偵測到錯誤之回應資訊(第三回應資訊(ReplyD(OK)或第四回應資訊(ReplyD(NG))傳送到記憶體控制器10。Furthermore, when the semiconductor memory device 20 receives the write data and the chip selection signal/CE is invalid (high level), it will send the response information (the third response information (ReplyD(OK) or the fourth response information (ReplyD(NG))) indicating whether an error is detected in the write data to the memory controller 10.

另外,半導體記憶裝置20在寫入資料被正確地寫入記憶胞陣列29內之記憶胞時,使訊號wloff有效(高位準)。因此,使訊號wl無效(低位準)。如此一來,記憶體控制器10從半導體記憶裝置20接收顯示在指令及位址中沒有偵測到錯誤之第一回應資訊(ReplyA(OK))時,與半導體記憶裝置20之間可以傳送對該位址寫入之資料。In addition, the semiconductor memory device 20 makes the signal wloff active (high level) when the write data is correctly written into the memory cells in the memory cell array 29 . Therefore, the signal wl is deasserted (low level). In this way, when the memory controller 10 receives the first response information (ReplyA (OK)) from the semiconductor memory device 20 indicating that no error has been detected in the command and address, it can transmit data written to the address with the semiconductor memory device 20 .

參照第3圖,以寫入指令從記憶體控制器10輸入的情況作為一例進行說明。Referring to FIG. 3 , a case where a write command is input from the memory controller 10 will be described as an example.

首先,半導體記憶裝置20從記憶體控制器10接收指令、位址及第一檢查資料時,利用第一檢查資料進行指令及位址之錯誤偵測。接著,半導體記憶裝置20在偵測到錯誤時,將顯示在指令及位址中之任一者偵測到錯誤的第二回應資訊(ReplyA(NG))傳送到記憶體控制器10。另外,在此情況下,半導體記憶裝置20不進行對應位址之字元線及位元線之活性化。Firstly, when the semiconductor memory device 20 receives the command, the address and the first check data from the memory controller 10, it uses the first check data to detect the error of the command and the address. Then, when an error is detected, the semiconductor memory device 20 transmits the second response information (ReplyA(NG)) indicating that any one of the command and address has detected an error to the memory controller 10 . In addition, in this case, the semiconductor memory device 20 does not activate the word line and the bit line corresponding to the address.

另一方面,記憶體控制器10在接收第二回應資訊(ReplyA(NG))時,將指令、位址及第一檢查資料再傳送到半導體記憶裝置20。On the other hand, when the memory controller 10 receives the second response information (ReplyA(NG)), it retransmits the command, the address and the first check data to the semiconductor memory device 20 .

半導體記憶裝置20從記憶體控制器10接收再傳送之指令、位址及第一檢查資料時,利用第一檢查資料進行指令 及位址之錯誤偵測。接著,半導體記憶裝置20在沒有偵測到錯誤時,將顯示在指令及位址中沒有偵測到錯誤之第一回應資訊(ReplyA(OK))傳送到記憶體控制器10。另外,此後半導體記憶裝置20之運作與第2圖所示之例子相同。When the semiconductor memory device 20 receives retransmitted commands, addresses and first check data from the memory controller 10, it uses the first check data to detect errors in the commands and addresses. Then, when no error is detected, the semiconductor memory device 20 sends the first response information (ReplyA(OK)) indicating that no error is detected in the command and address to the memory controller 10 . In addition, the operation of the semiconductor memory device 20 thereafter is the same as the example shown in FIG. 2 .

藉由在指令及位址中偵測到錯誤時,記憶體控制器10將指令及位址再傳送到半導體記憶裝置20,半導體記憶裝置20可以不開始錯誤之運作,從最初開始重新存取。When an error is detected in the command and address, the memory controller 10 retransmits the command and address to the semiconductor memory device 20, so that the semiconductor memory device 20 can resume access from the beginning without starting an error operation.

參照第4圖,以寫入指令從記憶體控制器10輸入的情況作為一例進行說明。Referring to FIG. 4, a case where a write command is input from the memory controller 10 will be described as an example.

首先,半導體記憶裝置20從記憶體控制器10接收指令、位址及第一檢查資料時,利用第一檢查資料進行指令及位址之錯誤偵測。另外,半導體記憶裝置20在晶片選擇訊號/CE從無效(高位準)變化為有效(低位準)時,使訊號busy有效(高位準)。Firstly, when the semiconductor memory device 20 receives the command, the address and the first check data from the memory controller 10, it uses the first check data to detect the error of the command and the address. In addition, the semiconductor memory device 20 makes the signal busy active (high level) when the chip select signal /CE changes from inactive (high level) to active (low level).

另外,半導體記憶裝置20在將第一回應資訊(ReplyA(OK))傳送到記憶體控制器10後,接收第二檢查資料及寫入資料時,利用第二檢查資料進行寫入資料之錯誤偵測。另外,如上所述,半導體記憶裝置20進行寫入資料之寫入。In addition, after the semiconductor memory device 20 transmits the first response information (ReplyA(OK)) to the memory controller 10, when receiving the second check data and write data, it uses the second check data to detect write data errors. In addition, as described above, the semiconductor memory device 20 performs writing of writing data.

此處,在寫入資料偵測到錯誤的情況下,半導體記憶裝置20在晶片選擇訊號/CE無效(高位準)時,將顯示在寫入資料中偵測到錯誤之第四回應資訊(ReplyD(NG))傳送到記憶體控制器10。另外,在此情況下,半導體記憶裝置20維持訊號wl及訊號busy各自有效之狀態。Here, when an error is detected in the written data, the semiconductor memory device 20 transmits the fourth response information (ReplyD(NG)) indicating that an error has been detected in the written data to the memory controller 10 when the chip select signal /CE is invalid (high level). In addition, in this case, the semiconductor memory device 20 maintains the respective valid states of the signal wl and the signal busy.

另一方面,記憶體控制器10接收第四回應資訊(ReplyD(NG))時,將第二檢查資料及寫入資料再傳送到半導體記憶裝置20。On the other hand, when the memory controller 10 receives the fourth response information (ReplyD(NG)), it retransmits the second check data and write data to the semiconductor memory device 20 .

半導體記憶裝置20從記憶體控制器10接收再傳送之第二檢查資料及寫入資料時,利用第二檢查資料進行寫入資料之錯誤偵測。另外,半導體記憶裝置20再度進行寫入資料之寫入。接著,在寫入資料沒有偵測到錯誤的情況下,半導體記憶裝置20將顯示在寫入資料中沒有偵測到錯誤之第三回應資訊(ReplyD(OK))傳送到記憶體控制器10。另外,半導體記憶裝置20使訊號wl及訊號busy各自無效(低位準)。When the semiconductor memory device 20 receives the retransmitted second check data and write data from the memory controller 10 , it uses the second check data to detect errors in the write data. In addition, the semiconductor memory device 20 performs writing of writing data again. Next, if no error is detected in the written data, the semiconductor memory device 20 transmits the third response information (ReplyD(OK)) indicating that no error has been detected in the written data to the memory controller 10 . In addition, the semiconductor memory device 20 disables (low level) the signal wl and the signal busy respectively.

如此一來,根據有關本實施型態之記憶體系統,可以抑制維持在讀取或寫入資料中包含錯誤的狀態。另外,根據有關本實施型態之記憶體系統,即使是在資料中偵測到錯誤的情況下,藉由等待該資料在傳送,半導體記憶裝置20可以繼續進行有關該資料之寫入的存取。另外,半導體記憶裝置20在等待資料再傳送的期間,藉由維持訊號wl有效的狀態,例如在指令及位址再傳送的情況下,因為變得不需要直到活性化感應放大器為止之等待時間(延遲),可以縮短運作時間。In this way, according to the memory system of this embodiment, it is possible to suppress the state in which errors are included in reading or writing data. In addition, according to the memory system related to the present embodiment, even in the case where an error is detected in the data, by waiting for the data to be transmitted, the semiconductor memory device 20 can continue access related to writing of the data. In addition, the semiconductor memory device 20 maintains the valid state of the signal wl while waiting for data to be retransmitted. For example, in the case of command and address retransmission, since the waiting time (delay) until the activation of the sense amplifier becomes unnecessary, the operation time can be shortened.

另外,在本實施型態中,雖然以在資料中偵測到錯誤的情況下,僅再傳送資料的情況作為一例進行說明,然而本發明不限於此情況。舉例而言,也可以再傳送指令 ,也可以再傳送對應列位址訊號ra及/或行位址訊號ca之位址,也可以再傳送指令及位址。In addition, in this embodiment, the case where only the data is retransmitted when an error is detected in the data is described as an example, but the present invention is not limited to this case. For example, it is also possible to retransmit the command, and also retransmit the address corresponding to the row address signal ra and/or the row address signal ca, or retransmit the command and the address.

再者,根據有關本實施型態之記憶體系統,在資料中偵測到錯誤的情況下,藉由記憶體控制器10再傳送資料,可以再度進行該資料之寫入。Furthermore, according to the memory system of this embodiment, when an error is detected in the data, the memory controller 10 retransmits the data, so that the data can be written again.

參照第5圖,以寫入指令從記憶體控制器10輸入的情況作為一例進行說明。Referring to FIG. 5, a case where a write command is input from the memory controller 10 will be described as an example.

首先,半導體記憶裝置20在寫入資料偵測到錯誤的情況下,使晶片選擇訊號/CE無效(高位準)時,將顯示在寫入資料中偵測到錯誤之第四回應資訊(ReplyD(NG))傳送到記憶體控制器10。另外,直到在寫入資料中偵測到錯誤為止的半導體記憶裝置20之運作,與參照第4圖說明之例子相同。First, when an error is detected in the written data, the semiconductor memory device 20 deactivates the chip selection signal /CE (high level), and transmits the fourth response information (ReplyD(NG)) indicating that an error has been detected in the written data to the memory controller 10 . In addition, the operation of the semiconductor memory device 20 until an error is detected in writing data is the same as the example described with reference to FIG. 4 .

此處,半導體記憶裝置20從第四回應資訊(ReplyD(NG))傳送到記憶體控制器10開始,直到經過特定時間為止的期間,在沒有從記憶體控制器10接收第二檢查資料及寫入資料的情況下,也可以使訊號wl及訊號busy各自無效(低位準)。因此,寫入資料之再傳送的待機狀態被解除。另外,半導體記憶裝置20也可以利用內藏之計時電路(圖示中省略)等測量經過時間。Here, the semiconductor memory device 20 may deactivate (low level) each of the signal wl and the signal busy when the second check data and write data are not received from the memory controller 10 during the period from the transmission of the fourth response information (ReplyD(NG)) to the memory controller 10 until a specific time elapses. Therefore, the standby state for retransmission of written data is released. In addition, the semiconductor memory device 20 may also measure elapsed time by using a built-in timer circuit (not shown in the figure) or the like.

如此一來,根據有關本實施型態之記憶體系統,半導體記憶裝置20在沒有再傳送資料的情況下,藉由解除等待該資料之再傳送,可以處理別的讀取或寫入存取。In this way, according to the memory system of the present embodiment, the semiconductor memory device 20 can process another read or write access by canceling the wait for the retransmission of the data without retransmitting the data.

接著,參照第6圖,半導體記憶裝置20在待機狀態(Standby)中,使晶片選擇訊號/CE有效(低位準),在指令、位址及第一檢查資料輸入時,利用第一檢查資料進行指令及位址之錯誤檢查(步驟S100)。接著,半導體記憶裝置20在偵測到錯誤的情況下(步驟S101:是),將第二回應資訊(ReplyA(NG)))輸出到記憶體控制器10,轉移到待機狀態。另外,半導體記憶裝置20在沒有偵測到錯誤的情況下(步驟S101:否),將第一回應資訊(ReplyA(OK))輸出到記憶體控制器10,轉移到活化狀態(Active)。Next, with reference to FIG. 6, the semiconductor memory device 20 in the standby state (Standby) enables the chip selection signal /CE to be valid (low level), and when the command, address and first check data are input, the first check data is used to perform command and address error checks (step S100). Next, when the semiconductor memory device 20 detects an error (step S101: Yes), it outputs the second response information (ReplyA(NG)) to the memory controller 10, and shifts to the standby state. In addition, when no error is detected (step S101: No), the semiconductor memory device 20 outputs the first response information (ReplyA (OK)) to the memory controller 10, and transitions to the active state (Active).

半導體記憶裝置20轉移到活化狀態(Active)時,進行記憶胞29之活性化(字元線以及位元線之活性化)(步驟S102)。When the semiconductor memory device 20 is transferred to the active state (Active), the memory cell 29 is activated (activation of the word line and the bit line) (step S102).

接著,半導體記憶裝置20根據指令及位址進行讀取或寫入存取(步驟S103)。此處,寫入存取的情況下,半導體記憶裝置20在接收第二檢查資料及寫入資料後,進行寫入存取。Next, the semiconductor memory device 20 performs read or write access according to the command and address (step S103 ). Here, in the case of write access, the semiconductor memory device 20 performs write access after receiving the second test data and write data.

另外,半導體記憶裝置20利用第二檢查資料進行寫入資料之錯誤檢查(步驟S104)。In addition, the semiconductor memory device 20 uses the second check data to check the error of the written data (step S104 ).

此處,在資料中偵測到錯誤時(步驟S105:是),半導體記憶裝置20半導體記憶裝置20將第四回應資訊(ReplyD(NG))輸出到記憶體控制器10,轉移到等待資料之再傳送的狀態。接著,晶片選擇訊號/CE為有效(低位準)時,半導體記憶裝置20轉移到步驟S103之處理。另外,半導體記憶裝置20在等待資料之再傳送的狀態中經過特定時間的情況下,轉移到後述之步驟S106的處理。在資料 中沒有偵測到錯誤時(步驟S105:否),半導體記憶裝置20將第三回應資訊(ReplyD(OK))輸出到記憶體控制器10,進行記憶胞陣列29之非活性化(字元線及位元線之非活性化)(步驟S106),轉移到待機狀態。Here, when an error is detected in the data (step S105: Yes), the semiconductor memory device 20 outputs the fourth response information (ReplyD(NG)) to the memory controller 10, and shifts to a state of waiting for retransmission of the data. Next, when the chip selection signal /CE is valid (low level), the semiconductor memory device 20 shifts to the processing of step S103. In addition, when the semiconductor memory device 20 waits for the retransmission of data for a certain period of time, it shifts to the processing of step S106 to be described later. When no error is detected in the data (step S105: No), the semiconductor memory device 20 outputs the third response information (ReplyD (OK)) to the memory controller 10, and performs inactivation of the memory cell array 29 (inactivation of the word line and the bit line) (step S106), and transfers to the standby state.

如上所述,根據有關本實施型態之記憶體系統,記憶體控制器10從半導體記憶裝置20接收顯示在指令及位址中沒有偵測到錯誤之第一回應資訊(ReplyA(OK))時,與半導體記憶裝置20之間可以傳送或接收對該位址讀取或寫入之資料。因此,指令包傳送中時,因為變得可以抑制由於指令之內容或位址變更導致進行不適當的存取,可以對半導體記憶裝置20適當地存取。另外,根據有關本實施型態之記憶體系統,以對虛擬靜態隨機存取記憶體適當地存取。As mentioned above, according to the memory system of this embodiment, when the memory controller 10 receives the first response information (ReplyA (OK)) from the semiconductor memory device 20 indicating that no error has been detected in the command and address, it can transmit or receive data read or written to the address with the semiconductor memory device 20. Therefore, when a command packet is being transmitted, it becomes possible to suppress inappropriate access due to a change in the content of the command or an address, and access to the semiconductor memory device 20 can be appropriately performed. In addition, according to the memory system of this embodiment, the virtual static random access memory can be properly accessed.

(第二實施型態) 以下說明本發明之第二實施型態之記憶體系統,在半導體記憶裝置20包括交錯式存取之複數個記憶庫。 (Second Implementation Type) The memory system of the second embodiment of the present invention will be described below. The semiconductor memory device 20 includes a plurality of memory banks for interleaved access.

參照第7圖及第8圖,,以半導體記憶裝置20包括2個記憶庫(BNK0及BNK1)的情況作為一例進行說明。第7(a)圖及第7(b) 圖顯示之時間圖,在第7圖(a)後接續第7(b)圖,顯示連續的時程,第8圖(a)、(b)亦同。Referring to FIG. 7 and FIG. 8 , the case where the semiconductor memory device 20 includes two memory banks (BNK0 and BNK1 ) will be described as an example. The time chart shown in Figure 7(a) and Figure 7(b) follows Figure 7(b) after Figure 7(a), showing a continuous time course, and the same is true for Figure 8(a) and (b).

如第7圖所示,半導體記憶裝置可以被配置為,從對記憶庫之一者(BNK0)輸入指令(Operation(BNK0))、位址(Address(BNK0))及第一檢查資料(CheckA(BNK0))開始,直到資料(Data(BNK0))輸入或輸出為止的延遲得期間,可以對記憶庫之另一者(BNK1)進行存取。As shown in FIG. 7, the semiconductor memory device can be configured such that the other memory bank (BNK1) can be accessed from the input of the command (Operation (BNK0)), the address (Address (BNK0)) and the first check data (CheckA (BNK0)) to one of the memory banks (BNK0) until the delay period until the data (Data (BNK0)) is input or output.

與第一實施型態同樣地,對記憶庫之一者(BNK0)在指令(Operation(BNK0))及位址(Address(BNK0))中沒有偵測到錯誤時,半導體記憶裝置20使對應位址(Address(BNK0,row=Rm(Rm顯示列位址,m為整數)))之字元線wl(BNK0)有效(高位準)。Same as the first embodiment, when no error is detected in the instruction (Operation (BNK0)) and address (Address (BNK0)) for one of the memory banks (BNK0), the semiconductor memory device 20 makes the word line wl (BNK0) of the corresponding address (Address (BNK0, row=Rm (Rm shows the column address, m is an integer))) effective (high level).

另外,半導體記憶裝置20對記憶庫之一者(BNK0)存取之延遲(延遲(BNK0))的期間,進行對記憶庫之另一者(BNK1)之存取。此處,半導體記憶裝置20在對記憶庫之另一者(BNK1)的存取中,進行資料(Data(BNK1))之讀取或寫入,該資料(Data(BNK1))對應對記憶庫之另一者(BNK1) 存取的延遲之前接收的位址(Address(BNK1,row=Ri(Ri顯示列位址,i為整數)))。接著,半導體記憶裝置20,在資料(Data(BNK1))中沒有偵測到錯誤時(ReplyD(BNK1)為OK時),接收對記憶庫之另一者(BNK1)的下一個存取指令(Operation(BNK1))及位址(Address(BNK1,row=Rj(Rj顯示列位址,j為整數)))。此時,半導體記憶裝置20使對應位址(Address(BNK1,row=Rj))之字元線wl(BNK1)有效(高位準)。In addition, the semiconductor memory device 20 performs access to the other memory bank (BNK1) during the delay (delay (BNK0)) of the access to one of the memory banks (BNK0). Here, the semiconductor memory device 20 reads or writes data (Data (BNK1)) corresponding to an address (Address (BNK1, row=Ri (Ri indicates a column address, i is an integer))) received before the delay in accessing the other memory bank (BNK1) during access to the other memory bank (BNK1). Next, the semiconductor memory device 20, when no error is detected in the data (Data (BNK1)) (ReplyD (BNK1) is OK), receives the next access command (Operation (BNK1)) and address (Address (BNK1, row=Rj (Rj displays the column address, j is an integer))) to the other memory bank (BNK1). At this time, the semiconductor memory device 20 makes the word line wl (BNK1) corresponding to the address (Address (BNK1, row=Rj)) active (high level).

接著,半導體記憶裝置20在對記憶庫之另一者(BNK1)存取之延遲(延遲(BNK1))的期間,對記憶庫之一者(BNK0)進行存取。此處,半導體記憶裝置20在對記憶庫之一者(BNK0)的存取中,進行資料(Data(BNK0))之讀取或寫入,該資料(Data(BNK0))對應對記憶庫之一者(BNK0) 存取的延遲之前接收的位址(Address(BNK0,row=Rm))。接著,半導體記憶裝置20,在資料(Data(BNK0))中沒有偵測到錯誤時(ReplyD(BNK0)為OK時),接收對記憶庫之一者(BNK0)的下一個存取指令(Operation(BNK0))及位址(Address(BNK0,row=Rn(Rn顯示列位址,n為整數)))。此時,半導體記憶裝置20使對應位址(Address(BNK0,row=Rn))之字元線wl(BNK0)有效(高位準)。Next, the semiconductor memory device 20 accesses one of the memory banks (BNK0) during the delay (delay (BNK1)) of access to the other memory bank (BNK1). Here, the semiconductor memory device 20 reads or writes data (Data (BNK0)) corresponding to an address (Address (BNK0, row=Rm)) received before the delay of access to one of the memory banks (BNK0) during access to one of the memory banks (BNK0). Next, the semiconductor memory device 20, when no error is detected in the data (Data (BNK0)) (ReplyD (BNK0) is OK), receives the next access command (Operation (BNK0)) and address (Address (BNK0, row=Rn (Rn shows the column address, n is an integer))) to one of the memory banks (BNK0). At this time, the semiconductor memory device 20 makes the word line wl (BNK0) corresponding to the address (Address (BNK0, row=Rn)) active (high level).

如第8圖所示,在對記憶庫之一者(BNK0)的存取中,在資料(Data(BNK0))中偵測到錯誤時(ReplyD(BNK0)為NG時),半導體記憶裝置20在接收沒有偵測到錯誤之資料(Data(BNK0))後,直到接收下一個存取指令(Operation(BNK0))及位址(Address(BNK0,row=Rn))為止,也可以延長對記憶庫之另一者(BNK1)的存取之延遲(延遲(BNK1))。As shown in FIG. 8, during access to one of the memory banks (BNK0), when an error is detected in the data (Data (BNK0)) (ReplyD (BNK0) is NG), after the semiconductor memory device 20 receives the data (Data (BNK0)) that has not detected an error, it can also receive the next access command (Operation (BNK0)) and address (Address (BNK0, row=Rn)). The delay of access to the other bank (BNK1) is extended (delay(BNK1)).

如上所述,根據本實施型態之記憶體系統,由於可以對複數個記憶庫同時並列進行存取,可以高速化對半導體記憶裝置20之存取。As described above, according to the memory system of the present embodiment, access to the semiconductor memory device 20 can be accelerated since a plurality of memory banks can be accessed in parallel at the same time.

(第三實施型態) 以下說明本發明之第三實施型態之記憶體系統,在半導體記憶裝置20被配置為在內部生成用以執行記憶胞之更新之更新要求。 (Third implementation type) A memory system according to a third embodiment of the present invention will be described below, in which the semiconductor memory device 20 is configured to internally generate refresh requests for performing memory cell refresh.

半導體記憶裝置20被配置為從生成更新要求訊號開始,直到執行更新為止之間,在從記憶體控制器10接收之指令、位址及資料中之任一者偵測到錯誤時,藉由從記憶體控制器10再傳送指令、位址及資料中之任一者,直到沒有偵測到錯誤為止,停止執行更新。The semiconductor memory device 20 is configured to stop updating by retransmitting any one of the command, address, and data from the memory controller 10 until no error is detected when an error is detected in any one of the command, address, and data received from the memory controller 10 between generation of the refresh request signal and execution of the refresh.

參照第9圖,以寫入指令從記憶體控制器10輸入的情況作為一例進行說明。Referring to FIG. 9, a case where a write command is input from the memory controller 10 will be described as an example.

首先,半導體記憶裝置20,在晶片選擇訊號/CE從有效(高位準)變化為無效(低位準)時,使訊號busy有效(高位準),半導體記憶裝置20在接收指令、位址及第一檢查資料的期間,在內部生成(有效)更新要求訊號refreq時,使用以等待更新之執行的訊號refwait有效(高位準)。舉例而言,更新要求訊號refreq及訊號refwait也可以由半導體記憶裝置20之指令控制部22生成。First, when the chip selection signal /CE changes from valid (high level) to invalid (low level), the semiconductor memory device 20 makes the signal busy valid (high level), and the semiconductor memory device 20 internally generates (valid) the update request signal refreq during the period of receiving instructions, addresses and first inspection data, and uses the signal refwait to wait for the execution of the update to be valid (high level). For example, the refresh request signal refreq and the signal refwait can also be generated by the command control unit 22 of the semiconductor memory device 20 .

接著,半導體記憶裝置20在指令及位址中偵測到錯誤時,將顯示偵測到錯誤的第二回應資訊(ReplyA(NG))傳送到記憶體控制器10。Next, when the semiconductor memory device 20 detects an error in the command and address, it sends the second response information (ReplyA(NG)) indicating the detected error to the memory controller 10 .

接著,半導體記憶裝置20在從記憶體控制器10再傳送之指令及位址 中沒有偵測到錯誤時,從記憶體控制器10接收寫入資料,進行接收之寫入資料的寫入。Then, when the semiconductor memory device 20 does not detect an error in the command and address retransmitted from the memory controller 10, it receives write data from the memory controller 10, and writes the received write data.

接著,半導體記憶裝置20在寫入資料 中沒有偵測到錯誤時 (第三回應資訊(ReplyD(OK))被傳送到記憶體控制器10時),使訊號refwait無效(低位準)。另外,半導體記憶裝置20在訊號refwait為無效(低位準)時執行更新。具體而言,半導體記憶裝置20為了活性化成為更新對象之字元線,使訊號wlon有效(高位準)的同時,使訊號wl有效(高位準)。另外,半導體記憶裝置20為了執行更新,使用以活性化位元線之訊號bl有效(高位準)。接著,半導體記憶裝置20執行更新。另外,半導體記憶裝置20在更新結束時,使訊號wloff有效(高位準)的同時,使訊號wl無效(低位準)。Next, when the semiconductor memory device 20 detects no error in the written data (when the third response information (ReplyD(OK)) is sent to the memory controller 10), the signal refwait is disabled (low level). In addition, the semiconductor memory device 20 executes refresh when the signal refwait is invalid (low level). Specifically, the semiconductor memory device 20 activates the signal wlon (high level) and the signal wl (high level) in order to activate the word line to be updated. In addition, the semiconductor memory device 20 uses the signal bl active (high level) for activating the bit line in order to execute refresh. Next, the semiconductor memory device 20 executes updating. In addition, when the refresh is completed, the semiconductor memory device 20 makes the signal wloff valid (high level) and at the same time makes the signal wl invalid (low level).

雖然在圖中省略說明,半導體記憶裝置20在資料中偵測到錯誤時,也可以等待更新之執行(訊號refwait為有效(高位準)),直到變得沒有從記憶體控制器10傳送之資料偵測到錯誤為止。Although the description is omitted in the figure, when the semiconductor memory device 20 detects an error in the data, it can also wait for the execution of the update (the signal refwait is valid (high level)), until no error is detected in the data transmitted from the memory controller 10.

藉由直到沒有在指令、位址及資料中之任一者偵測到錯誤為止,停止更新之執行,可以抑制起因於執行更新之存取延遲。By stopping the execution of the update until no error is detected in any of the command, address, and data, it is possible to suppress the access delay caused by performing the update.

另外,作為第三實施型態之變形例,半導體記憶裝置20從生成更新要求訊號refreq開始,直到執行更新為止的期間,也可以在從記憶體控制器10接收之指令、位址及資料中之任一者偵測到錯誤時,將顯示執行更新之第五回應資訊傳送到記憶體控制器10,執行更新。In addition, as a modified example of the third embodiment, the semiconductor memory device 20 may also transmit the fifth response information indicating the execution of the update to the memory controller 10 when an error is detected in any one of the command, address, and data received from the memory controller 10 during the period from generating the update request signal refreq to executing the update, and performing the update.

參照第10圖,以寫入指令從記憶體控制器10輸入的情況作為一例進行說明。Referring to FIG. 10, a case where a write command is input from the memory controller 10 will be described as an example.

首先,半導體記憶裝置20在晶片選擇訊號/CE從無效(高位準)變化為有效(低位準)時,使訊號busy有效(高位準)。此處,半導體記憶裝置20在接收指令 、位址及第一檢查資料的期間,在內部生成(有效)更新要求訊號refreq時,使用以等待更新之執行的訊號refwait有效。First, the semiconductor memory device 20 makes the signal busy active (high level) when the chip selection signal /CE changes from inactive (high level) to active (low level). Here, when the semiconductor memory device 20 internally generates (validates) the refresh request signal refreq during the period of receiving the command, the address and the first check data, the signal refwait used to wait for the execution of the refresh is valid.

接著,半導體記憶裝置20在指令及位址偵測到錯誤時,將顯示偵測到錯誤的第二回應資訊(ReplyA(NG))傳送到記憶體控制器10。此處,也可以在第二回應資訊(ReplyA(NG))中包含顯示執行更新之第五回應資訊(在圖中顯示為”Ref next”)。Next, the semiconductor memory device 20 transmits the second response information (ReplyA(NG)) indicating the detected error to the memory controller 10 when an error is detected in the command and address. Here, the second response information (ReplyA(NG)) may also include fifth response information (shown as "Ref next" in the figure) indicating the update execution.

另一方面,記憶體控制器10在接收包含第五回應資訊(”Ref next”)之第二回應資訊(ReplyA(NG))時,也可以考慮在半導體記憶裝置20內執行更新,直到經過特定時間為止,等待下一個存取中的命令(Operation) 、位址及第一檢查資料之傳送。On the other hand, when the memory controller 10 receives the second response information (ReplyA(NG)) including the fifth response information ("Ref next"), it may also consider performing an update in the semiconductor memory device 20 until a certain time elapses, waiting for the transmission of the next accessing command (Operation), address and first check data.

接著,半導體記憶裝置20使訊號busy無效(低位準)一次,其後使其有效(高位準)。另外,半導體記憶裝置20,回應無效(低位準)之訊號busy開始執行更新,使訊號refwait無效(低位準)。接著,半導體記憶裝置20在結束執行更新後,進行下一個存取中之處理。Next, the semiconductor memory device 20 deactivates the signal busy (low level) once, and then enables it (high level). In addition, the semiconductor memory device 20 responds to the invalid (low level) signal busy to start updating, and makes the signal refwait invalid (low level). Next, the semiconductor memory device 20 performs the processing in the next access after finishing execution of the update.

藉由在指令、位址及資料中之任一者偵測到錯誤時執行更新,舉例而言,可以抑制反覆執行指令、位址及資料中之任一者之再傳送的期間,起因於沒有執行更新之半導體記憶裝置20的記憶資訊遺失。By executing refresh when an error is detected in any of command, address, and data, for example, it is possible to suppress loss of memory information of semiconductor memory device 20 due to non-refreshing during a period in which retransmission of any of command, address, and data is repeatedly performed.

上述實施型態中,雖然以半導體記憶裝置20為pSRAM的情況作為一例進行說明,本發明不限於此情況。舉例而言,半導體記憶裝置20也可以是DRAM,也可以是快閃記憶體,也可以是其他半導體記憶裝置。In the above-mentioned embodiments, although the case where the semiconductor memory device 20 is a pSRAM is taken as an example for description, the present invention is not limited to this case. For example, the semiconductor memory device 20 may also be a DRAM, a flash memory, or other semiconductor memory devices.

另外,上述實施型態中,雖然以記憶體控制器10將寫入指令(Operation(Write))傳送到半導體記憶裝置20的情況作為一例進行說明,即使是記憶體控制器10將讀取指令(Operation(Read))傳送到半導體記憶裝置20的情況,也可以得到與上述實施型態同樣的作用效果。In addition, in the above-mentioned embodiment, although the case where the memory controller 10 transmits the write command (Operation (Write)) to the semiconductor memory device 20 is described as an example, even if the memory controller 10 transmits the read command (Operation (Read)) to the semiconductor memory device 20, the same effect as the above-mentioned embodiment can be obtained.

半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之一者),也可以將讀取資料及讀取資料之錯誤偵測用之第二檢查資料傳送到記憶體控制器10(記憶體控制器10及半導體記憶裝置20中之另一者)。另外,記憶體控制器10接收讀取資料及讀取資料之錯誤偵測用之第二檢查資料時,也可以利用第二檢查資料進行讀取資料之錯誤偵測,在讀取資料中沒有偵測到錯誤時,將顯示在讀取資料中沒有偵測到錯誤之第三回應資訊(ReplyD(OK))傳送到半導體記憶裝置20,可以確認讀取資料已適當地到達記憶體控制器10。The semiconductor memory device 20 (one of the memory controller 10 and the semiconductor memory device 20) may also transmit the read data and the second check data for error detection of the read data to the memory controller 10 (the other of the memory controller 10 and the semiconductor memory device 20). In addition, when the memory controller 10 receives the read data and the second check data for error detection of the read data, it can also use the second check data to detect errors in the read data. When no error is detected in the read data, the third response information (ReplyD (OK)) indicating that no error has been detected in the read data is sent to the semiconductor memory device 20 to confirm that the read data has properly reached the memory controller 10.

另外,記憶體控制器10(記憶體控制器10及半導體記憶裝置20中之另一者)在資料中偵測到錯誤時,也可以將顯示在資料中偵測到錯誤之第四回應資訊(ReplyD(NG))傳送到半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之一者),直到經過特定時間為止等待從半導體記憶裝置20再傳送資料。即使是在資料中偵測到錯誤時,記憶體控制器10藉由等待該資料再傳送,可以繼續進行有關該資料之讀取之存取。因此,與從最初開始重新存取有關該資料之讀取的情況(再傳送指令及位址的情況)比較之下,可以提高存取效率。In addition, when the memory controller 10 (the other of the memory controller 10 and the semiconductor memory device 20) detects an error in the data, it can also transmit the fourth response information (ReplyD(NG)) indicating that an error has been detected in the data to the semiconductor memory device 20 (one of the memory controller 10 and the semiconductor memory device 20), and wait until a specific time passes to transmit the data from the semiconductor memory device 20. Even when an error is detected in the data, the memory controller 10 can continue the access related to the read of the data by waiting for the data to be retransmitted. Therefore, the access efficiency can be improved compared with the case of re-accessing the reading of the data from the beginning (the case of re-sending the command and the address).

再者還有,半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之一者)在接收第四回應資訊(ReplyD(NG))時,也可以將資料及對應資料之第二檢查資料再傳送到記憶體控制器10(記憶體控制器10及半導體記憶裝置20中之另一者)。因此,在資料中偵測到錯誤時,藉由半導體記憶裝置20再傳送資料,可以再度進行該資料之讀取。Furthermore, when the semiconductor memory device 20 (one of the memory controller 10 and the semiconductor memory device 20) receives the fourth response information (ReplyD(NG)), it can also send the data and the second check data corresponding to the data to the memory controller 10 (the other one of the memory controller 10 and the semiconductor memory device 20). Therefore, when an error is detected in the data, the data can be read again by retransmitting the data through the semiconductor memory device 20 .

另外,記憶體控制器(記憶體控制器10及半導體記憶裝置20中之另一者)直到經過特定時間為止,在沒有從半導體記憶裝置20(記憶體控制器10及半導體記憶裝置20中之一者)再傳送資料的情況下,也可以解除等待資料之再傳送。在此情況下,記憶體控制器10在沒有再傳送資料的情況下,藉由解除等待該資料之再傳送,可以處理別的讀取或寫入存取,可以提升存取之處理效率。In addition, the memory controller (the other one of the memory controller 10 and the semiconductor memory device 20) may cancel waiting for retransmission of data when no data is retransmitted from the semiconductor memory device 20 (one of the memory controller 10 and the semiconductor memory device 20) until a specific time elapses. In this case, the memory controller 10 can process other read or write accesses by canceling the waiting for the retransmission of the data without retransmitting the data, which can improve the processing efficiency of the access.

上述實施型態中,雖然以第一檢查資料與指令及位址以分開的狀態被傳送的情況作為一例進行說明,本發明不限於此情況。在第11圖中,顯示指令 及位址之位元分割之一例。另外,此處假設半導體記憶裝置20為pSRAM的情況。In the above-mentioned embodiment, although the case where the first inspection data is transmitted separately from the command and the address is described as an example, the present invention is not limited to this case. Fig. 11 shows an example of bit division of commands and addresses. In addition, it is assumed here that the semiconductor memory device 20 is a pSRAM.

在第11圖所示的例子中,外部時脈訊號CLK之每個上升邊緣及下降邊緣輸入8位元之位址資料訊號ADQ[7]~ADQ[0]。指令被配置在第1個外部時脈訊號CLK之上升邊緣輸入的位址資料訊號ADQ[7]~ADQ[5]中,位址(Address[31]~Address[0])被配置在從第1個外部時脈訊號CLK到第3個外部時脈訊號CLK中被輸入之位址資料訊號ADQ中。另外,雖然在第3個外部時脈訊號CLK之上升邊緣中輸入的位址資料訊號ADQ[7]~ADQ[0],以及在第3個外部時脈訊號CLK之下降邊緣中輸入的位址資料訊號ADQ[7]~ADQ[3]為已預約(Reserved)之區域,卻是未使用之區域。此處,也可以利用該預約區域之至少一部分,配置為第一檢查資料。因此,半導體記憶裝置20變得可以同時接收第一檢查資料與位址。另外,第一檢查資料也可以被配置在第4個以後的外部時脈訊號CLK之上升邊緣或下降邊緣中輸入之位址資料訊號ADQ中。In the example shown in FIG. 11 , each rising edge and falling edge of the external clock signal CLK inputs 8-bit address data signals ADQ[7]˜ADQ[0]. The command is configured in the address data signal ADQ[7]~ADQ[5] input on the rising edge of the first external clock signal CLK, and the address (Address[31]~Address[0]) is configured in the address data signal ADQ input from the first external clock signal CLK to the third external clock signal CLK. In addition, although the address data signals ADQ[7]~ADQ[0] input during the rising edge of the third external clock signal CLK and the address data signals ADQ[7]~ADQ[3] input during the falling edge of the third external clock signal CLK are reserved areas, they are unused areas. Here, at least a part of the reserved area may be used as the first inspection data. Therefore, the semiconductor memory device 20 becomes able to receive the first inspection data and the address at the same time. In addition, the first inspection data may also be arranged in the address data signal ADQ input on the rising edge or falling edge of the fourth and subsequent external clock signal CLK.

另外,資料選通訊號RWDS也可以被配置以代表顯示執行更新之第五回應資訊。舉例而言,資料選通訊號RWDS為高位準(H)時,可以代表顯示執行更新之第五回應資訊(“Ref next”)。另一方面,資料選通訊號RWDS為低位準(L)時,也可以代表顯示可以直接傳送接收位址資料訊號ADQ 之訊號(“Immediate”)。In addition, the data strobe signal RWDS can also be configured to represent the fifth response information for display execution update. For example, when the data strobe signal RWDS is at a high level (H), it may represent the fifth response message (“Ref next”) for displaying and updating. On the other hand, when the data strobe signal RWDS is at a low level (L), it can also represent a signal (“Immediate”) that can directly transmit and receive the address data signal ADQ.

如第12圖所示,資料選通訊號RWDS也可以被配置為代表第一回應資訊(ReplyA(OK))或第二回應資訊(ReplyA(NG))。舉例而言,第4個外部時脈訊號CLK的上升邊緣中,維持資料選通訊號RWDS之邏輯位準(高位準或低位準)時(“kept”),資料選通訊號RWDS也可以代表第一回應資訊(ReplyA(OK))。另一方面,第4個外部時脈訊號CLK的上升邊緣中,反轉資料選通訊號RWDS之邏輯位準(高位準或低位準)時(“inverted”),資料選通訊號RWDS也可以代表第二回應資訊(ReplyA(NG))。另外,舉例而言,也可以回應在第5個外部時脈訊號CLK之上升邊緣或下降邊緣中之資料選通訊號RWDS之狀態,代表第一回應資訊(ReplyA(OK))或第二回應資訊(ReplyA(NG))。As shown in FIG. 12, the data strobe signal RWDS can also be configured to represent the first response message (ReplyA(OK)) or the second response message (ReplyA(NG)). For example, when the logic level (high level or low level) of the data strobe signal RWDS is maintained (“kept”) during the rising edge of the fourth external clock signal CLK, the data strobe signal RWDS can also represent the first response message (ReplyA(OK)). On the other hand, when the logic level (high level or low level) of the data strobe signal RWDS is inverted (“inverted”) on the rising edge of the fourth external clock signal CLK, the data strobe signal RWDS can also represent the second response information (ReplyA(NG)). In addition, for example, it is also possible to respond to the state of the data strobe signal RWDS on the rising edge or falling edge of the fifth external clock signal CLK, representing the first response information (ReplyA(OK)) or the second response information (ReplyA(NG)).

10: 記憶體控制器 11: 要求控制部 12: 序列控制部 12a: 串聯器/解串器(Serializer/deserializer)(SerDes) 12b: 錯誤控制部 20: 半導體記憶裝置 21: 序列控制部 21a: 串聯器/解串器(SerDes) 21b: 錯誤控制部 22: 指令控制部 23: 位址控制部 24: 字元線控制部 25: 行控制部 26: 資料控制部 27: 資料匯流排控制部 28: 感應放大器 29: 記憶胞陣列 Address:位址 ADQ: 位址資料訊號 BANK、BNK0、BNK1: 記憶庫 /CE: 晶片選擇訊號 CheckA: 第一檢查資料 CheckD: 第二檢查資料 CLK: 外部時脈訊號 Data: 資料 Deson、Seron: 變換訊號 Operation: 指令 ReplyA: 回應資訊 ReplyD: 回應資訊 Retryrd、Retrywr: 訊號 RWDS: 資料選通訊號 S100~S108: 流程圖步驟 adr、bl、busy: 訊號 ca: 行位址訊號 cdb、rdb、wdb: 訊號 cl: 位元線 clon: 活性化位元線之訊號 di、do:訊號 ope: 訊號 ra: 列位址訊號 refreq: 更新要求訊號 refwait: 訊號 saon: 訊號 wl: 字元線 wloff: 訊號 wlon: 訊號 10: Memory Controller 11: Request Control Department 12: Sequence Control Department 12a: Serializer/deserializer (Serializer/deserializer) (SerDes) 12b: Error control section 20: Semiconductor memory device 21: Sequence Control Department 21a: Serializer/Deserializer (SerDes) 21b: Error Control Section 22: Command Control Department 23: Address Control Section 24: Character Line Control Section 25: Row Control Department 26: Data Control Department 27: Data bus control section 28: Sense amplifier 29: Memory cell array Address: address ADQ: Address data signal BANK, BNK0, BNK1: memory banks /CE: chip select signal CheckA: first check data CheckD: The second check data CLK: external clock signal Data: data Deson, Seron: Change the signal Operation: instruction ReplyA: Response information ReplyD: Response information Retryrd, Retrywr: Signal RWDS: data strobe signal S100~S108: Flow chart steps adr, bl, busy: signal ca: row address signal cdb, rdb, wdb: signal cl: bit line clon: signal to activate the bit line di, do: signal ope: signal ra: row address signal refreq: Refresh request signal refwait: signal saon: signal wl: character line wloff: signal wlon: signal

[第1圖]為顯示有關本發明第一實施型態之記憶體系統之構成例的方塊圖。 [第2圖]為顯示在指令及位址中沒有包含錯誤時的半導體記憶裝置內之訊號之時程的時間圖。 [第3圖]為顯示在指令及位址的任一者中包含錯誤時的半導體記憶裝置內之訊號之時程的時間圖。 [第4圖]為顯示資料中包含錯誤時的半導體記憶裝置內之訊號之時程的時間圖。 [第5圖]為顯示等待資料再傳送時經過特定時間的情況下,半導體記憶裝置內之訊號之時程的時間圖。 [第6圖]為顯示半導體記憶裝置之運作之一例的流程圖。 [第7圖](a)、(b)為顯示在有關本發明第二實施型態之記憶體系統中,半導體記憶裝置內之訊號之時程的時間圖。 [第8圖](a)、(b)為顯示半導體記憶裝置內之訊號之時程的時間圖。 [第9圖] 為顯示在有關本發明第三實施型態之記憶體系統中,半導體記憶裝置內之訊號之時程的時間圖。 [第10圖] 為顯示半導體記憶裝置內之訊號之時程的時間圖。 [第11圖]為顯示指令及位址的位元分配之一例的示意圖。 [第12圖]為顯示資料選通訊號之時程的時間圖。 [FIG. 1] is a block diagram showing a configuration example of a memory system related to the first embodiment of the present invention. [Fig. 2] is a time chart showing the timing of signals in the semiconductor memory device when no errors are included in commands and addresses. [FIG. 3] is a time chart showing the timing of signals in the semiconductor memory device when an error is included in any of the command and the address. [Fig. 4] is a time chart showing the timing of signals in a semiconductor memory device when data contains errors. [FIG. 5] is a time chart showing the time course of signals in a semiconductor memory device when a specific time elapses while waiting for data to be retransmitted. [FIG. 6] is a flow chart showing an example of the operation of a semiconductor memory device. [Fig. 7] (a) and (b) are time charts showing the timing of signals in the semiconductor memory device in the memory system related to the second embodiment of the present invention. [Fig. 8] (a) and (b) are time charts showing the timing of signals in the semiconductor memory device. [FIG. 9] It is a time chart showing the timing of signals in the semiconductor memory device in the memory system related to the third embodiment of the present invention. [Fig. 10] It is a time chart showing the timing of signals in the semiconductor memory device. [FIG. 11] is a schematic diagram showing an example of bit allocation of commands and addresses. [FIG. 12] is a timing diagram showing the timing of the data strobe signal.

10: 記憶體控制器 11: 要求控制部 12: 序列控制部 12a: 串聯器/解串器(Serializer/deserializer)(SerDes) 12b: 錯誤控制部 20: 半導體記憶裝置 21: 序列控制部 21a: 串聯器/解串器(SerDes) 21b: 錯誤控制部 22: 指令控制部 23: 位址控制部 24: 字元線控制部 25: 行控制部 26: 資料控制部 27: 資料匯流排控制部 28: 感應放大器 29: 記憶胞陣列 Address:位址 ADQ: 位址資料訊號 /CE: 晶片選擇訊號 CheckA: 第一檢查資料 CheckD: 第二檢查資料 CLK: 外部時脈訊號 Data: 資料 Deson、Seron: 變換訊號 Operation: 指令 ReplyA: 回應資訊 ReplyD: 回應資訊 Retryrd、Retrywr: 訊號 RWDS: 資料選通訊號 adr、bl、busy: 訊號 ca: 行位址訊號 cdb、rdb、wdb: 訊號 cl: 位元線 clon: 活性化位元線之訊號 di、do:訊號 ope: 訊號 ra: 列位址訊號 saon: 訊號 wl: 字元線 wloff: 訊號 wlon: 訊號 10: Memory Controller 11: Request Control Department 12: Sequence Control Department 12a: Serializer/deserializer (Serializer/deserializer) (SerDes) 12b: Error control section 20: Semiconductor memory device 21: Sequence Control Department 21a: Serializer/Deserializer (SerDes) 21b: Error Control Section 22: Command Control Department 23: Address Control Section 24: Character Line Control Section 25: Row Control Department 26: Data Control Department 27: Data bus control section 28: Sense amplifier 29: Memory cell array Address: address ADQ: Address data signal /CE: chip select signal CheckA: first check data CheckD: The second check data CLK: external clock signal Data: data Deson, Seron: Change the signal Operation: instruction ReplyA: Response information ReplyD: Response information Retryrd, Retrywr: Signal RWDS: data strobe signal adr, bl, busy: signal ca: row address signal cdb, rdb, wdb: signal cl: bit line clon: signal to activate the bit line di, do: signal ope: signal ra: row address signal saon: signal wl: word line wloff: signal wlon: signal

Claims (15)

一種記憶體系統,包括: 記憶體控制器;以及 半導體記憶裝置; 前述記憶體控制器,被配置為進行: 將指令及位址,以及前述指令及前述位址之錯誤偵測用之第一檢查資料傳送到前述半導體記憶裝置;以及 從前述半導體記憶裝置接收顯示在前述指令及前述位址沒有偵測到錯誤之第一回應資訊時,與前述半導體記憶裝置之間,根據前述指令傳送或接收對前述位址讀取或寫入之資料; 前述半導體記憶裝置,被配置為進行: 從前述記憶體控制器接收前述指令及前述位址,以及前述第一檢查資料時,利用前述第一檢查資料進行前述指令及前述位址之錯誤偵測,在前述指令及前述位址沒有偵測到錯誤時,將前述第一回應資訊傳送到前述記憶體控制器;以及 在前述指令及前述位址沒有偵測到錯誤時,與前述記憶體控制器之間,根據前述指令傳送或接收對前述位址讀取或寫入之資料。 A memory system comprising: memory controller; and Semiconductor memory devices; The aforementioned memory controller, configured to: transmitting the command and address, and the first check data for error detection of the aforementioned command and the aforementioned address to the aforementioned semiconductor memory device; and When receiving the first response information from the aforementioned semiconductor memory device indicating that no error has been detected in the aforementioned command and the aforementioned address, transmit or receive data read or written to the aforementioned address according to the aforementioned command with the aforementioned semiconductor memory device; The aforementioned semiconductor memory device configured to: When receiving the aforementioned command, the aforementioned address, and the aforementioned first check data from the aforementioned memory controller, use the aforementioned first check data to perform error detection on the aforementioned command and the aforementioned address, and when no error is detected in the aforementioned command and the aforementioned address, send the aforementioned first response information to the aforementioned memory controller; and When no error is detected in the aforementioned command and the aforementioned address, the data read or written to the aforementioned address is transmitted or received according to the aforementioned command with the aforementioned memory controller. 如請求項1之記憶體系統,其中,前述半導體記憶裝置被配置為: 在前述指令及前述位址之任一者偵測到錯誤時,將顯示在前述指令及前述位址之任一者偵測到錯誤的第二回應資訊傳送到前述記憶體控制器; 前述記憶體控制器被配置為: 從前述半導體記憶裝置接收前述第二回應資訊時,將前述指令及前述位址再傳送到前述半導體記憶裝置。 The memory system according to claim 1, wherein the aforementioned semiconductor memory device is configured as: When any one of the aforementioned command and the aforementioned address detects an error, sending the second response information indicating that any of the aforementioned command and the aforementioned address detects an error to the aforementioned memory controller; The foregoing memory controller is configured as: When receiving the second response information from the semiconductor memory device, retransmit the command and the address to the semiconductor memory device. 如請求項1之記憶體系統,其中 前述記憶體控制器及前述半導體記憶裝置中之一者,被配置為: 將前述讀取或寫入之資料,以及前述資料之錯誤偵測用之第二檢查資料,傳送到前述記憶體控制器及前述半導體記憶裝置中之另一者; 前述記憶體控制器及前述半導體記憶裝置中之另一者,被配置為: 接收前述資料以及前述第二檢查資料時,利用前述第二檢查資料進行前述資料之錯誤偵測,在前述資料中沒有偵測到錯誤時,將顯示沒有在前述資料中偵測到錯誤的第三回應資訊傳送到前述記憶體控制器及前述半導體記憶裝置中之一者。 Such as the memory system of claim 1, wherein One of the aforementioned memory controller and the aforementioned semiconductor memory device is configured to: Sending the aforementioned read or written data, and the second check data for error detection of the aforementioned data to the other of the aforementioned memory controller and the aforementioned semiconductor memory device; The other of the aforementioned memory controller and the aforementioned semiconductor memory device is configured to: When receiving the aforementioned data and the aforementioned second check data, use the aforementioned second check data to perform error detection on the aforementioned data, and when no error is detected in the aforementioned data, send the third response information indicating that no error has been detected in the aforementioned data to one of the aforementioned memory controller and the aforementioned semiconductor memory device. 如請求項3之記憶體系統,其中 前述記憶體控制器及前述半導體記憶裝置中之另一者,被配置為進行: 在前述資料中偵測到錯誤時,將顯示在前述資料中偵測到錯誤的第四回應資訊,傳送到前述記憶體控制器及前述半導體記憶裝置中之一者;以及 直到經過特定時間為止,等待從前述記憶體控制器及前述半導體記憶裝置中之一者再傳送前述資料。 Such as the memory system of claim 3, wherein The other of the aforementioned memory controller and the aforementioned semiconductor memory device is configured to perform: When an error is detected in the aforementioned data, a fourth response message indicating that an error is detected in the aforementioned data is sent to one of the aforementioned memory controller and the aforementioned semiconductor memory device; and Waiting for the data to be retransmitted from one of the memory controller and the semiconductor memory device until a specific time elapses. 如請求項4之記憶體系統,其中 前述記憶體控制器及前述半導體記憶裝置中之一者,被配置為: 接收前述第四回應資訊時,將前述資料以及對應前述資料之前述第二檢查資料,再傳送到前述記憶體控制器及前述半導體記憶裝置中之另一者。 Such as the memory system of claim 4, wherein One of the aforementioned memory controller and the aforementioned semiconductor memory device is configured to: When receiving the aforementioned fourth response information, the aforementioned data and the aforementioned second inspection data corresponding to the aforementioned data are sent to the other of the aforementioned memory controller and the aforementioned semiconductor memory device. 如請求項4之記憶體系統,其中 前述記憶體控制器及前述半導體記憶裝置中之另一者,被配置為: 直到經過前述特定時間為止,沒有從前述記憶體控制器及前述半導體記憶裝置中之一者再傳送前述資料時,解除等待前述資料之再傳送。 Such as the memory system of claim 4, wherein The other of the aforementioned memory controller and the aforementioned semiconductor memory device is configured to: When the data is not retransmitted from one of the memory controller and the semiconductor memory device until the specified time elapses, the waiting for retransmission of the data is cancelled. 如請求項1之記憶體系統,其中,前述半導體記憶裝置更包括:以交錯方式被存取之複數個記憶庫(bank)。The memory system according to claim 1, wherein the aforementioned semiconductor memory device further includes: a plurality of memory banks (banks) accessed in an interleaved manner. 如請求項1之記憶體系統,其中,在前述半導體記憶裝置被配置為在內部生成用以執行更新之更新要求訊號的情況下,從生成前述更新要求訊號開始,到前述更新被執行為止之間,從前述記憶體控制器接收之前述指令、前述位址及前述資料中之任一者偵測到錯誤時,藉由從前述記憶體控制器再傳送前述指令、前述位址及前述資料中之任一者,直到前述錯誤變得沒有被偵測到為止,停止前述更新之執行。The memory system according to claim 1, wherein, when the semiconductor memory device is configured to internally generate an update request signal for performing the update, when an error is detected in any of the command, the address, and the data received from the memory controller between the generation of the refresh request signal and the execution of the update, execution of the update is stopped by retransmitting any of the command, address, and data from the memory controller until the error becomes undetected. 如請求項1之記憶體系統,其中,在前述半導體記憶裝置被配置為在內部生成用以執行更新之更新要求訊號的情況下,被配置為從生成前述更新要求訊號開始,到前述更新被執行為止之間,從前述記憶體控制器接收之前述指令、前述位址及前述資料中之任一者偵測到錯誤時,將顯示執行前述更新之第五回應資訊傳送到前述記憶體控制器,執行前述更新。The memory system according to claim 1, wherein, when the semiconductor memory device is configured to internally generate an update request signal for performing the update, it is configured to transmit fifth response information indicating that the update is executed to the memory controller when an error is detected in any of the command, the address, and the data received from the memory controller between the generation of the update request signal and the execution of the update. 如請求項1之記憶體系統,其中,前述半導體記憶裝置為虛擬靜態隨機存取記憶體。The memory system according to claim 1, wherein the aforementioned semiconductor memory device is a virtual static random access memory. 如請求項1之記憶體系統,其中前述記憶體控制器,包括: 要求控制部,從主機裝置接收對前述半導體記憶裝置寫入或讀取之要求時,生成變換訊號,前述變換訊號用以在前述記憶體控制器與前述半導體記憶裝置之間變換傳送或接收之位址資料訊號的傳送方式;以及 第一序列控制部,基於前述變換訊號變換前述位址資料訊號的傳送方式; 其中 前述第一序列控制部包括: 第一串聯器/解串器,基於前述變換訊號將前述位址訊號串列化或反串列化;以及 第一錯誤控制部,利用前述指令以及前述位址生成前述第一檢查資料; 其中,前述第一串聯器/解串器被配置為進行: 在前述要求控制部接收寫入的情況下,前述變換訊號從前述要求控制部輸入時,生成寫入指令以及位址輸出到前述第一錯誤控制部; 對應生成之寫入指令以及位址的第一檢查資料從前述第一錯誤控制部輸入時,將生成之寫入指令以及位址和前述第一檢查資料作為前述位址資料訊號傳送到前述半導體記憶裝置;以及 在從前述半導體記憶裝置接收顯示生成之寫入指令以及位址中沒有偵測到錯誤的第一回應資訊作為前述位址資料訊號時,將從前述主機裝置接收的寫入資料輸出到前述第一錯誤控制部。 The memory system according to claim 1, wherein the aforementioned memory controller includes: The request control unit, when receiving a request from the host device to write or read the semiconductor memory device, generates a conversion signal, the conversion signal is used to convert the transmission method of the address data signal transmitted or received between the memory controller and the semiconductor memory device; and The first sequence control unit converts the transmission method of the address data signal based on the conversion signal; in The aforementioned first sequence control unit includes: a first serializer/deserializer serializing or deserializing the address signal based on the conversion signal; and The first error control unit uses the aforementioned instruction and the aforementioned address to generate the aforementioned first inspection data; Wherein, the aforementioned first serializer/deserializer is configured to: When the request control unit receives write, when the conversion signal is input from the request control unit, a write command and an address are generated and output to the first error control unit; When the first check data corresponding to the generated write command and address is input from the first error control unit, the generated write command and address and the first check data are sent to the semiconductor memory device as the address data signal; and When receiving the generated write command from the semiconductor memory device and the first response information indicating that no error has been detected in the address as the address data signal, the write data received from the host device is output to the first error control unit. 如請求項11之記憶體系統,其中,前述第一錯誤控制部被配置為利用前述寫入資料,生成前述寫入資料的錯誤偵測用之第二檢查資料; 前述串聯器/解串器被配置為在前述第二檢查資料從前述第一錯誤控制部輸入時,將前述第二檢查資料以及前述寫入資料作為前述位址資料訊號傳送到前述半導體記憶裝置。 The memory system according to claim 11, wherein the first error control unit is configured to use the written data to generate second check data for error detection of the written data; The serializer/deserializer is configured to transmit the second check data and the write data as the address data signal to the semiconductor memory device when the second check data is input from the first error control unit. 如請求項11之記憶體系統,其中,前述第一錯誤控制部被配置為從前述半導體記憶裝置接收顯示從前述寫入指令以及位址中的任一者偵測到錯誤的第二回應資訊作為前述位址資料訊號時,生成用以再傳送前述寫入指令以及位址和前述第一檢查資料的訊號; 前述第一串聯器/解串器被配置為在從前述半導體記憶裝置接收前述第二回應資訊作為前述位址資料訊號的情況下,在用以再傳送前述寫入指令以及位址和前述第一檢查資料的訊號從前述第一錯誤控制部輸入時,將前述寫入指令以及位址和前述第一檢查資料作為前述位址資料訊號再傳送到前述半導體記憶裝置。 The memory system according to claim 11, wherein the first error control unit is configured to generate a signal for retransmitting the write command and address and the first check data when receiving second response information indicating that an error has been detected from any of the write command and address from the semiconductor memory device as the address data signal; The first serializer/deserializer is configured to, when receiving the second response information from the semiconductor memory device as the address data signal, retransmit the write command, the address and the first check data as the address data signal to the semiconductor memory device when a signal for retransmitting the write command, the address, and the first check data is input from the first error control section. 如請求項11之記憶體系統,其中,前述半導體記憶裝置更包括第二序列控制部,變換與前述記憶體控制器之間傳送或接收的前述位址資料訊號的傳送方式; 前述第二序列控制部包括: 第二串聯器/解串器,將前述位址資料訊號串列化或反串列化;以及 第二錯誤控制部,利用從前述記憶體控制器接收之前述第一檢查資料進行前述指令以及前述位址的錯誤偵測; 前述第二串聯器/解串器被配置為進行: 在從前述記憶體控制器接收前述指令以及前述位址和前述第一檢查資料作為前述位址資料訊號的情況下,將前述指令以及前述位址和前述第一檢查資料變換為串列傳輸方式輸出到前述第二錯誤控制部; 前述第一回應資訊從前述第二錯誤控制部輸入時,將前述第一回應資訊變換為並列傳輸方式,將前述第一回應資訊作為前述位址資料訊號輸出到記憶體控制器,將顯示前述指令之內容的訊號輸出到基於前述指令生成內部指令的指令控制部,將顯示前述位址的訊號,輸出到控制對應前述位址的字元線以及位元線活性化的位址控制部;以及 前述第二回應資訊從前述第二錯誤控制部輸入時,將前述第二回應資訊變換為並列傳輸方式,前述第二回應資訊作為前述位址資料訊號傳送到前述記憶體控制器。 The memory system according to claim 11, wherein the aforementioned semiconductor memory device further includes a second serial control unit for converting the transmission method of the aforementioned address data signal transmitted or received between the aforementioned memory controller; The aforementioned second sequence control unit includes: a second serializer/deserializer for serializing or deserializing the aforementioned address data signal; and The second error control unit uses the first inspection data received from the memory controller to detect errors of the command and the address; The aforementioned second serializer/deserializer is configured to: In the case of receiving the aforementioned command, the aforementioned address, and the aforementioned first check data from the aforementioned memory controller as the aforementioned address data signal, converting the aforementioned command, the aforementioned address, and the aforementioned first check data into a serial transmission method and outputting it to the aforementioned second error control unit; When the aforementioned first response information is input from the aforementioned second error control unit, the aforementioned first response information is converted into a parallel transmission method, the aforementioned first response information is output to the memory controller as the aforementioned address data signal, the signal displaying the content of the aforementioned command is output to the command control unit that generates an internal command based on the aforementioned command, and the signal displaying the aforementioned address is output to the address control unit that controls activation of the word line and bit line corresponding to the aforementioned address; and When the aforementioned second response information is input from the aforementioned second error control unit, the aforementioned second response information is transformed into a parallel transmission method, and the aforementioned second response information is sent to the aforementioned memory controller as the aforementioned address data signal. 如請求項14之記憶體系統,其中 前述第二錯誤控制部被配置為利用前述寫入資料之錯誤偵測用之第二檢查資料進行前述寫入資料之錯誤偵測; 前述第二串聯器/解串器被配置為進行: 從前述記憶體控制器接收前述寫入資料以及前述寫入資料的錯誤偵測用之第二檢查資料作為位址資料訊號時,將前述寫入資料以及前述第二檢查資料輸出到前述第二錯誤控制部; 在顯示在前述寫入資料中偵測到錯誤的第四回應資訊從第二錯誤控制部輸入時,將前述第四回應資訊作為前述位址資料訊號傳送到前述記憶體控制器。 Such as the memory system of claim 14, wherein The aforementioned second error control unit is configured to use the second check data for error detection of the aforementioned written data to perform error detection of the aforementioned written data; The aforementioned second serializer/deserializer is configured to: When receiving the write data and second check data for error detection of the write data from the memory controller as address data signals, output the write data and the second check data to the second error control unit; When the fourth response information indicating that an error is detected in the write data is input from the second error control unit, the fourth response information is sent to the memory controller as the address data signal.
TW110147330A 2021-12-17 2021-12-17 Memory system TWI807542B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW110147330A TWI807542B (en) 2021-12-17 2021-12-17 Memory system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW110147330A TWI807542B (en) 2021-12-17 2021-12-17 Memory system

Publications (2)

Publication Number Publication Date
TWI807542B true TWI807542B (en) 2023-07-01
TW202326742A TW202326742A (en) 2023-07-01

Family

ID=88147724

Family Applications (1)

Application Number Title Priority Date Filing Date
TW110147330A TWI807542B (en) 2021-12-17 2021-12-17 Memory system

Country Status (1)

Country Link
TW (1) TWI807542B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI459391B (en) * 2009-11-06 2014-11-01 Toshiba Kk Memory system
CN105895164A (en) * 2015-02-16 2016-08-24 力晶科技股份有限公司 Semiconductor Memory Apparatus And Semiconductor Integrated Circuit Apparatus
TWI640986B (en) * 2017-03-09 2018-11-11 東芝記憶體股份有限公司 Semiconductor memory device and data reading method
US20190096470A1 (en) * 2009-01-20 2019-03-28 Longitude Licensing Limited Semiconductor memory device, method of controlling read preamble signal thereof, and data transmission system
TW202137230A (en) * 2020-03-17 2021-10-01 日商鎧俠股份有限公司 Semiconductor device and semiconductor storage device
TW202147118A (en) * 2020-06-02 2021-12-16 日商鎧俠股份有限公司 Memory system and memory controller

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190096470A1 (en) * 2009-01-20 2019-03-28 Longitude Licensing Limited Semiconductor memory device, method of controlling read preamble signal thereof, and data transmission system
TWI459391B (en) * 2009-11-06 2014-11-01 Toshiba Kk Memory system
CN105895164A (en) * 2015-02-16 2016-08-24 力晶科技股份有限公司 Semiconductor Memory Apparatus And Semiconductor Integrated Circuit Apparatus
TWI640986B (en) * 2017-03-09 2018-11-11 東芝記憶體股份有限公司 Semiconductor memory device and data reading method
TW202137230A (en) * 2020-03-17 2021-10-01 日商鎧俠股份有限公司 Semiconductor device and semiconductor storage device
TW202147118A (en) * 2020-06-02 2021-12-16 日商鎧俠股份有限公司 Memory system and memory controller

Also Published As

Publication number Publication date
TW202326742A (en) 2023-07-01

Similar Documents

Publication Publication Date Title
JP6986369B2 (en) Memory module, system including it and its operation method
JP6815898B2 (en) Memory modules, memory systems, and methods
TWI567748B (en) Method and system for error management in a memory device
US8687456B2 (en) Multi-port memory based on DRAM core
US8286054B2 (en) Semiconductor memory, operating method of semiconductor memory, and system
KR101048380B1 (en) Memory module device
US10957413B2 (en) Shared error check and correct logic for multiple data banks
JP2002366444A (en) System and method for correcting soft error in random access memory device
JP2011227948A (en) Semiconductor memory device and control method thereof
JP2011081553A (en) Information processing system and control method thereof
US20130086449A1 (en) Sharing a Check Bit Memory Device Between Groups of Memory Devices
US7239569B2 (en) Semiconductor memory device and memory system
TWI807542B (en) Memory system
JP7428689B2 (en) memory system
KR102645215B1 (en) Memory system
JP2001266570A (en) Synchronous semiconductor memory
JP5107776B2 (en) Memory controller, memory system, and method of writing data to memory device
CN116610471A (en) Memory system
KR20170128783A (en) Memory system and operation method of the same
JP2005267354A (en) Semiconductor device
JP2002197864A (en) Multi-port memory and its control method
JP5200914B2 (en) Semiconductor memory and system
JP7500698B2 (en) Circuit and method for reading ECC from memory - Patents.com
KR20190102530A (en) Address generting circuit, address and command generating circuit and semiconductor system
JP2009217310A (en) Memory access method and memory access device