TW425575B - Address space manager for assigning part of address space to data storage without non-addresable cell, semiconductor memory device with built-in address space manager and memory system - Google Patents

Address space manager for assigning part of address space to data storage without non-addresable cell, semiconductor memory device with built-in address space manager and memory system Download PDF

Info

Publication number
TW425575B
TW425575B TW088108696A TW88108696A TW425575B TW 425575 B TW425575 B TW 425575B TW 088108696 A TW088108696 A TW 088108696A TW 88108696 A TW88108696 A TW 88108696A TW 425575 B TW425575 B TW 425575B
Authority
TW
Taiwan
Prior art keywords
address
patent application
address space
memory device
semiconductor memory
Prior art date
Application number
TW088108696A
Other languages
Chinese (zh)
Inventor
Takemi Kimura
Original Assignee
Nippon Electric Co
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 Nippon Electric Co filed Critical Nippon Electric Co
Application granted granted Critical
Publication of TW425575B publication Critical patent/TW425575B/en

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/06Addressing a physical block of locations, e.g. base addressing, module addressing, memory dedication
    • G06F12/0646Configuration or reconfiguration
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02Addressing or allocation; Relocation
    • G06F12/08Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G06F12/10Address translation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601Interfaces specially adapted for storage systems
    • G06F3/0628Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0629Configuration or reconfiguration of storage systems
    • G06F3/0631Configuration or reconfiguration of storage systems by allocating resources to storage systems
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C8/00Arrangements for selecting an address in a digital store
    • G11C8/10Decoders

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Read Only Memory (AREA)
  • Dram (AREA)
  • Memory System (AREA)

Abstract

An address space manager (31c/31d) assigns address sub-spaces forming parts of an address space to memory units (31a/31b) at least one (31b) of which stores m (6M) pieces of data information where m is greater than (n-1)<SP>th</SP> power of 2 and less than n<SP>th</SP> power of 2, wherein the address space manger (31c/31d) converts external address bits (A21/A22) to internal address bits (A22'/A22') in such a manner that the addresses are respectively assigned to the memory cells of the memory units (31a/31b) without non-addressable memory cell.

Description

L 4 2 5 5 7 5 ' 五、發明說明u) 【發明領域】 本發明係有關於位址空間管理技術,尤其有關於用以 指定部份位址空間至資料儲存器例如半導體記憶裝置之位 址空間管理器,具有内建式位址空間管理器與記憶系統之 半導體記憶裝置。 【相關技術之說明】 半導體記憶裝置之典型例子係揭露於日本專利公開公 報第6 2 - 5 2 7 9 4號中。先前技術半導體記憶裝置係歸類於唯 ..讀記憶裝置,並且先前技術半導體唯讀記憶裝置包含記憶 單元陣列與關於記憶單元陣列之位址解碼器。記憶單元陣 列包含複數個記憶單元,並且複數個記憶單元形成位址空 間。輸入位址信號係代表大於記憶單元陣列之位址空間之 位址空間,並且供應至位址解碼器。程式構件進一步地併 入先前技術半導體唯讀記憶裝置中,並且使位址解碼信號 之組合轉換成位址解碼信號之另一組合。使用者可以程式 化此轉換,並且自由地指定由輸入位址信號所代表之部份 位址空間至記憶單元陣列。 圖1係闡明先前技術記憶系統1之配置。微處理器2伴 隨有先前技術記憶系統1。先前技術記憶系統1包含半導體 記憶裝置1 a / 1 b / 1 c / 1 d與位址解碼器1 e,並且每個半導 體記憶裝置la - 1 d係假設成具有用以儲存二個資料數元組 Μ 0 / Μ 1之記憶單元陣列1 f n g / 1 h / 1 j。微處理器2使位址信 號A 0與位址信號A 1 / A 2分別供應至半導體記憶裝置L 4 2 5 5 7 5 'V. Description of the Invention u) [Field of the Invention] The present invention relates to address space management technology, and more particularly, to designating a portion of the address space to a data storage device such as a semiconductor memory device. Address space manager, a semiconductor memory device with a built-in address space manager and a memory system. [Explanation of Related Technology] A typical example of a semiconductor memory device is disclosed in Japanese Patent Laid-Open Publication No. 6 2-5 2 7 9 4. The prior art semiconductor memory device is classified as a read-only memory device, and the prior art semiconductor read-only memory device includes a memory cell array and an address decoder for the memory cell array. The memory cell array includes a plurality of memory cells, and the plurality of memory cells form an address space. The input address signal represents an address space larger than the address space of the memory cell array and is supplied to the address decoder. The program component is further incorporated into a prior art semiconductor read-only memory device, and converts a combination of address decoded signals into another combination of address decoded signals. The user can program this conversion and freely specify a portion of the address space represented by the input address signal to the memory cell array. FIG. 1 illustrates the configuration of a prior art memory system 1. The microprocessor 2 is accompanied by a prior art memory system 1. The prior art memory system 1 includes a semiconductor memory device 1 a / 1 b / 1 c / 1 d and an address decoder 1 e, and each semiconductor memory device 1-1 d is assumed to have two data elements for storing two data. Memory cell array of group M 0 / M 1 1 fng / 1 h / 1 j. The microprocessor 2 supplies the address signal A 0 and the address signal A 1 / A 2 to the semiconductor memory device, respectively.

第5頁 1 425575 ^_____ ________ _ — 五、發明說明(2) la/lb/lc/ld之位址輸入埠lk/lm/ln/lo與位址解碼器le ° 位址信號A 0係表示指定至用以儲存二種類型Μ 0 / Μ 1中之一 之記憶單元之位址。另一方面,位址信號A 1 / A 2係表示半 導體記憶裝置la-ld中之一。 位址解碼器le包含反相器IV1/IV2與NAND閘 ND1/ND2/ND3/ND4。反相器IV1/IV2從位址信號A1/A2分別 產生反相位址信號A B 1 / A B 2,並且位址信號A 1 / A 2與反相位 址信號A B 1 / A B 2形成四種組合。位址信號A 1 / A 2與反相位址 信號A B 1 / A B 2之四種組合係分別供應至N A N D閘N D 1 - N D 4,並 且NAND閘ND1-ND4從四種組合產生解碼信號 DE1/DE2/DE3/DE4。四種組合中僅一種具有二個輸入皆為 邏輯&quot;1”位準’並且相關聯的N A N D閑使解碼信號改變為低 態有效位準。 N A N D閘N D 1 - N D 4係分別連接至半導體記憶裝置1 a - 1 d之 晶片選擇/輸出賦能埠0E1/0E2/0E3/0E4。低態有效位準之 解碼信號使半導體記憶裝置1 a - 1 d中相關聯之一個回應於 位址彳s號A 0,並且任一類型Μ 0 / Μ 1係從半導體記憶裝置之 資料輸出埠〇1]11/011丁2/01]丁3/01^4讀至微處理器2^中。雖然 半導aa 5己憶裝置具有其他控制信號插針例如寫入賦能插 針,但是圖1所示之半導體記憶裝置i a_丨d中則省略之。 圖2闡明位址信號A0/A1/A2與受存取資料位元間 之關係。位址信號A 2 = 0與A 1 = 0使位址解碼器1 e啟動半_ 導體記憶裝置1 a,半導體記憶裝置1 a依據位址信號a q之邏 輯位準使類型MO/Ml可存取。當位址信號A2/A1分別為低位Page 5 1 425575 ^ _____ ________ _ — V. Description of the invention (2) The address input port lk / lm / ln / lo of la / lb / lc / ld and the address decoder le ° address signal A 0 indicates The address assigned to the memory unit used to store one of the two types of M 0 / M 1. On the other hand, the address signal A 1 / A 2 represents one of the semiconductor memory devices la-ld. The address decoder le includes inverters IV1 / IV2 and NAND gates ND1 / ND2 / ND3 / ND4. The inverters IV1 / IV2 generate inverted phase address signals AB 1 / AB 2 from the address signals A1 / A2, respectively, and the address signals A 1 / A 2 and the inverted phase address signals AB 1 / AB 2 form four combinations. . The four combinations of the address signals A 1 / A 2 and the inverted phase address signals AB 1 / AB 2 are supplied to the NAND gates ND 1-ND 4 respectively, and the NAND gates ND1-ND4 generate decoded signals DE1 / from the four combinations. DE2 / DE3 / DE4. Only one of the four combinations has two inputs that are both logic &quot; 1 &quot; and the associated NAND idler causes the decoded signal to change to an active low level. NAND gates ND 1-ND 4 are connected to semiconductor memory respectively Chip select / output enable port 0E1 / 0E2 / 0E3 / 0E4 of device 1 a-1 d. The low-level effective level decoded signal causes the associated one of semiconductor memory devices 1 a-1 d to respond to address 彳 s No. A 0, and any type Μ 0 / Μ 1 is read from the data output port of the semiconductor memory device 〇1] 11/011 丁 2/01] 丁 3/01 ^ 4 into the microprocessor 2 ^. Although half The lead aa 5 device has other control signal pins such as write enable pins, but the semiconductor memory device i a_ 丨 d shown in Figure 1 is omitted. Figure 2 illustrates the address signal A0 / A1 / A2 Relation to the bit of the data being accessed. The address signals A 2 = 0 and A 1 = 0 cause the address decoder 1 e to activate the semi-conductor memory device 1 a, and the semiconductor memory device 1 a according to the address signal aq. Logic level makes type MO / Ml accessible. When address signals A2 / A1 are low

五、發明說明(3) 準與高位準時,半導體記憶裝置1 b會回應於位址信號A 0, 並且資料數元組Μ 0 / Μ 1係依據位址信號A 0之邏輯位準而選 擇性地從半導體記憶裝置1 b中讀出。假若位址信號A 2 / A 1 分別改變成高位準與低位準時,則位址解碼器1 e會啟動半 導體記憶裝置1 c,並且資料數元組Μ 0 / Μ 1係依據位址信號 Α0之邏輯位準而選擇性地存取。最後,當位址信號Α2/Α1 二者皆為高位準時,位址解碼器1 e會選擇半導體記憶裝置 1 d,並且資料數元組Μ 0 / Μ 1會依據位址信號A 0之邏輯位準 而選擇性地從半導體記憶裝置1 d中讀出。 曰本專利公開公報第6 2 - 5 2 7 9 4號所揭露之先前技術半 導體唯讀記憶裝置係等同於任一具有相當於位址解碼器1 e 之内建式位址解碼器之半導體記憶裝置1 a / 1 b / 1 c / 1 d。 半導體記憶裝置之貢料儲存容夏在近十年之間業已劇 烈地增加,據此,位址空間變廣。半導體記憶裝置係使用 於各種電/電子系統中,並且在此等電]»/電子系統間所需之 資料儲存容量係不同的。製造業者業已以2的η次方增加先 前技術半導體記憶裝置之資料儲存容4。舉例而言,製造 業者業已使半導體記憶裝置之資料儲存容量增加超過3 2千 數元組與64千數元組而達128千數元組。表示為2的η次方 之資料儲存器經常適合於一般目的之資料儲存。程控指令 碼之集合係儲存於先前技術半導體唯讀記憶裝置中,並且 程控指令碼之集合得佔用4 8千數元組之記憶空間,而另一 程控指令碼之集合得佔用9 6千數元組之記憶空間。此等程 控指令碼之集合有時使用於電腦遊戲中。未表示為2的η次V. Description of the invention (3) On-time and high-time, the semiconductor memory device 1 b will respond to the address signal A 0, and the data tuple M 0 / M 1 is selective according to the logic level of the address signal A 0 Ground is read from the semiconductor memory device 1b. If the address signal A 2 / A 1 is changed to the high level and the low level, respectively, the address decoder 1 e will start the semiconductor memory device 1 c, and the data byte M 0 / M 1 is based on the address signal A0. Logical level and selective access. Finally, when the address signals A2 / A1 are both high level, the address decoder 1e will select the semiconductor memory device 1d, and the data byte M0 / M1 will be based on the logical bit of the address signal A0 Read from the semiconductor memory device 1 d accurately and selectively. The prior art semiconductor read-only memory device disclosed in this Patent Publication No. 6 2-5 2 7 9 4 is equivalent to any semiconductor memory having a built-in address decoder equivalent to the address decoder 1 e Device 1 a / 1 b / 1 c / 1 d. The memory storage capacity of semiconductor memory devices has increased dramatically in the past ten years, and accordingly, the address space has become wider. Semiconductor memory devices are used in various electric / electronic systems, and the data storage capacity required between these electric] »/ electronic systems is different. Manufacturers have increased the data storage capacity of prior art semiconductor memory devices by a factor of two. For example, manufacturers have increased the data storage capacity of semiconductor memory devices by more than 32 kilobytes and 64 kilobytes to 128 kilobytes. Data stores denoted as powers of n are often suitable for general purpose data storage. The set of program-controlled scripts is stored in the semiconductor read-only memory device of the prior art, and the set of program-controlled scripts can occupy 4 to 8 thousand bytes of memory space, while the set of another program-controlled script can occupy 9 to 6 thousand Group memory space. This collection of program control scripts is sometimes used in computer games. Not represented η times

42557 5 五、發明說明(4) 方之資料儲存容量在下文中係稱為「2的η次方以下之資料 儲存容量」。當位址信號供應於半導體記憶裝置之η個位 址插針上時,半導體記憶裝置會提供表示為2η之位址空 間。假若在相同的環境下另一半導體記憶裝置具有介於2η 與2nM間之資料儲存容量,則半導體記憶裝置會提供稱為 「2的η次方以下之資料儲存容量」之位址空間。 電腦設計者可利用不同記憶空間之半導體唯讀記憶裝 置以組織2的η次方以下之記憶空間。另外,電腦設計者得 使用具有記憶空間大於所需記憶空間之半導體唯讀記憶裝 置。舉例言之,倘若電腦設計者在他的程控指令碼中需要 9 6千數元組之記憶空間,則他得採用具有6 4千數元組之記 憶空間之半導體唯讀記憶裝置與具有3 2千數元組之記憶空 間之半導體唯讀記憶裝置。另外,他得使用具有1 2 8千數 元組之記憶空間之半導體唯讀記憶裝置。無論如何,此二 種程控指令碼之記憶系統皆不經濟。相似地,當記憶單元 陣列之一部份有缺陷時,因為設計者係基於表示為2的η次 方之資料儲存容量而組織位址空間,所以先前技術半導體 記憶裝置無法用作為資料儲存器。 曰本專利公開公報中揭露之先前技術半導體唯讀記憶 裝置係提供表示為2的η次方之記憶空間。即使程式構件使 位址解碼信號之組合轉換成位址解碼信號之另一組合,位 址解碼信號仍僅管理表示為2的η次方之記憶空間,並且曰 本專利公開公報未採用2的η次方以下之資料儲存容量。 當電腦設計者一起管理具有2的η次方以下之資料儲存42557 5 V. Description of the invention (4) The data storage capacity of Party F is hereinafter referred to as "data storage capacity below 2 to the nth power". When the address signal is supplied to n address pins of the semiconductor memory device, the semiconductor memory device provides an address space expressed as 2η. If another semiconductor memory device has a data storage capacity between 2n and 2nM under the same environment, the semiconductor memory device will provide an address space called "data storage capacity below the power of n". Computer designers can use semiconductor read-only memory devices with different memory spaces to organize memory spaces below the nth power. In addition, computer designers may use semiconductor read-only memory devices that have more memory space than needed. For example, if a computer designer needs 9 6 thousand tuples of memory space in his program-controlled instruction code, he must use a semiconductor read-only memory device with a 6 4 thousand tuples memory space and a 3 2 Semiconductor read-only memory device with thousands of tuples of memory space. In addition, he had to use a semiconductor read-only memory device with a memory space of 128 bytes. In any case, the memory system of these two program-controlled scripts is not economical. Similarly, when a part of the memory cell array is defective, the prior art semiconductor memory device cannot be used as a data storage because the designer organizes the address space based on the data storage capacity represented by the power of n expressed as 2. The prior art semiconductor read-only memory device disclosed in this patent publication provides a memory space denoted by a power of n. Even if the program component converts the combination of the address decoded signal into another combination of the address decoded signal, the address decoded signal still only manages the memory space represented by the power of n to the power of 2 and said that this patent publication does not adopt the power of 2 Data storage capacity below the power. When a computer designer manages a data store with a power of 2n or less

7 4 2 5 5 7 5 ’ 五、發明說明(5) 容量之半導體δ己憶裝置之s己憶空間以及具有表示為2的n次 方之資料儲存容量之另一半導體記憶裝置之記憶空間時, 如下文所述s己憶糸統會具有不受任何位址指定之部分位址 空間及/或受位址多重指定之部分位址空間。 圖3係闡明先前技術記憶系統1 〇。先前技術記憶系統 10包含一個半導體5己憶裝置l〇a/10b與位址解碼琴iQc'/單 一位址位元A 0供應至半導體記憶裝置1 q &amp;之位址插針a 〇, 並且半導體記憶裝置1 0 a提供關於二個資料數元組M 1之 記憶空間。因此,半導體記憶裝置1 0 a之資料储存容量係 表示為21數元組’並且半導體記憶裝置i〇a係具有表示為2 的π次方之資料健存谷里之類型。另一方面,三個位址位 元A 0 / A 1 / A 2係供應至半導體記憶裝置1 〇 b之位址插針 a0/al/a2 ’並且半導體記憶裝置提供關於六個資料數元組 M0/M1/M2/M3/M4/M5之記憶空間。半導體記憶裝置i〇b之資 料儲存容董係表示為(22+2),其小於23。因此,半導體 記憶裝置1 0 b係具有表示為2的η次方以下之資料儲存容量 之類型。半導體記憶裝置l〇a/l〇b分別具有資料輸出埠 0UT1/0UT2 *並且連接至共用資料匯流排BSi。 位址解瑪器1 〇 C具有〇 r閘〗0 d與反相器1 〇 e。位址位元 AJ / A 2,係供應至〇 R閘1 〇 d之輪入節點,並且〇 R閘丨〇 d之輸出 節點係使輸出賦能/晶片選擇信號供應至半導體記憶裝置 l〇a之輸出賦能/晶片選擇插針〇E1與反 點。反相器!,,輸出賦能/晶片選擇信號益產°生, 並且使反相#號供應至半導體記憶裝置丨〇b之輸出賦能/晶7 4 2 5 5 7 5 'V. Description of the invention (5) The memory space of the semiconductor delta memory device of the capacity and the memory space of another semiconductor memory device having a data storage capacity of the power of n expressed as 2. As described below, the system will have a portion of the address space that is not designated by any address and / or a portion of the address space that is designated by multiple addresses. Figure 3 illustrates a prior art memory system 10. The prior art memory system 10 includes a semiconductor memory device 10a / 10b and an address decoder iQc '/ single address bit A 0 supplied to the address pins a 0 of the semiconductor memory device 1 q &amp; and The semiconductor memory device 10a provides a memory space for two data tuples M1. Therefore, the data storage capacity of the semiconductor memory device 10 a is represented as 21-tuples', and the semiconductor memory device i 0a has a type of data that is represented by a power of π. On the other hand, the three address bits A 0 / A 1 / A 2 are supplied to the address pins a0 / al / a2 'of the semiconductor memory device 10b and the semiconductor memory device provides about six data bytes M0 / M1 / M2 / M3 / M4 / M5 memory space. The data storage capacity of the semiconductor memory device i0b is expressed as (22 + 2), which is less than 23. Therefore, the semiconductor memory device 10b is a type having a data storage capacity expressed as a power of n or less. The semiconductor memory devices 10a / 10b respectively have data output ports OUT1 / 0UT2 * and are connected to a common data bus BSi. The address resolver 10 C has 0 r gate 0 d and inverter 10 e. The address bit AJ / A2 is supplied to the turn-on node of the gate IOd, and the output node of the gate 〇0d enables the output enable / chip selection signal to be supplied to the semiconductor memory device 10a Output enable / chip selection pin OE1 and anti-point. Inverter !, the output enable / wafer select signal is generated, and the inverse # is supplied to the semiconductor memory device. The output enable / crystal

第9頁 ^ 425575 五、發明說明(6) 片選擇插針0E2。 當位址位元A 1 / A 2二者皆位於邏輯” 〇 &quot;位準時,〇 R閘 1 0 d使輸出賦能/晶片選擇信號保持在低態有效位準,並且 半導體記憶裝置10a會回應於位址位元A〇。另一方面,當 位址位元A1/A2中至少一個在高位準時,閘}0d會使輸出 賦能/晶片選擇信號改變成高態無效位準,並且反相器1 0 e 會使另一半導體記憶裝置1 0 b回應於位址位元A 0 / A 1 / A 2。 雖然在先前技術記憶系統1 〇中之記憶單元係等於8資 料數元組,但是記憶系統1 0僅提供關於6資料數元組之位 址空間。如圖4中所示位址位元A 2、A 1 、A 0係依序地增 加。當位址位元A2、A1與A0係[〇, 0, 0 ]與[〇, 〇, 1 ]時,半導 體記憶裝置1 0 a中之資料數元組Μ 0與Μ 1會變成可存取。當 外部裝置使位址位元Α 2 / A 1 / A 0從[0,1,0 ]增加至[1,〇,1 ] 時,資料數元組M2、M3、M4與M5會依序變成可存取。位址 位元[0,〇,0 ]與[0,0,1 ]之組合業已指定至半導體記憶裝置 1 0 a。因此,位址位元A 2、A 1與A 0無法從半導體記憶裝置 1 0 b中選擇資料數元組Μ 〇 ϋ之記憶單元。 圖5係闡明另一先前技術記憶系統2 0。先前技術記慎 系統2 0包含二個半導體記憶裝置2 〇 a / 2 〇 b與位址解碼器 c ‘。雖然一個位址位元A 0 / A 1係供應至半導體記憶裝置 20a之位址插針a0/al ’但是半導體記憶裝置20a提供'關於 三個資料數元組Μ 〇 / Μ 1 / Μ 2之§己憶空間。因此’半導體吃慎 裝置2 0a之資料儲存容量係表示為(22-1 )個數元組,並心 且半導體記憶裝置2 0 a係具有2的η次方以下之資料儲存容Page 9 ^ 425575 V. Description of the invention (6) Pin selection pin 0E2. When the address bits A 1 / A 2 are both at the logic level “0”, the gate 1 0 d keeps the output enable / chip selection signal at a low valid level, and the semiconductor memory device 10a will In response to address bit A0. On the other hand, when at least one of address bits A1 / A2 is at a high level, the gate} 0d will change the output enable / chip selection signal to a high state invalid level and reverse The phaser 10 e will cause another semiconductor memory device 10 b to respond to the address bits A 0 / A 1 / A 2. Although the memory cell in the prior art memory system 10 is equal to 8 data bytes, But memory system 10 only provides address space for 6 data bytes. As shown in Figure 4, address bits A2, A1, A0 are sequentially increased. When address bits A2, A1 When A0 is [〇, 0, 0] and [〇, 〇, 1], the data bytes M 0 and M 1 in the semiconductor memory device 10 a become accessible. When the external device makes the address bit When elements A 2 / A 1 / A 0 increase from [0,1,0] to [1, 〇, 1], the data tuples M2, M3, M4, and M5 will become accessible sequentially. Address bits Yuan [0, 〇, 0] The combination of [0, 0, 1] has been assigned to the semiconductor memory device 10a. Therefore, the address bits A2, A1, and A0 cannot select the data byte group M0 from the semiconductor memory device 10b Figure 5 illustrates another prior art memory system 20. The prior art caution system 20 includes two semiconductor memory devices 20a / 2b and an address decoder c '. Although one address bit The elements A 0 / A 1 are supplied to the address pins a0 / al of the semiconductor memory device 20 a ′ but the semiconductor memory device 20 a provides ′ a memory space for three data tuples M 0 / M 1 / M 2. Therefore, the data storage capacity of 'semiconductor device 2 0a' is expressed as (22-1) number of tuples, and the semiconductor memory device 2 0a has a data storage capacity of 2 to the power of η or less.

4 2 5 5 7 5 五、發明說明(7) 量之類型。另一方面,二個位址位元A 〇 / A 1係供應至半導 體記憶裝置20b之位址插針a〇/ai ,半導體記憶裝置2〇1^提 供關於四個資料數元組Μ 4 / Μ 5 / Μ 6 / Μ 7之記憶空間。半導體 記憶裝置2 0 b之資料儲存容量係表示為。因此,半導體 記憶裝置20b係具有表示為2的η次方之資料儲存容量之類 型。半導體記憶裝置2 0 a / 2 0 b分別具有資料輸出埠 0 U T 1 / 0 U T 2 ’並且連接至共用資料匯流排B s !。 位址解碼器2 0 c具有反相器2 0 d。位址位元A 2係供應至 半導體記憶裝置2 0 a之輸出賦能/晶片選擇插針〇E 1以作為 輸出賦能/晶片選擇信號。反相器2 0 d產生輸出賦能/晶片 選擇信號之反相信號,並且使反相信號供應至輸出賦能/ 晶片選擇插針0E 2。因此,位址解碼器2 〇 c使半導體記憶裝 置20a/20b中之一回應於位址位元A0/A1。 如圖6中所示位址位元A2、A1與A0係依序地改變。當 位址位元A 2係邏輯” 〇”位準時,位址解碼器2 〇 c會使半導體 記憶裝置2 0 a回應於位址位元A 0 / A 1。另一方面,邏輯1 &quot; 位準之位址位元A 2允許另一半導體記憶裝置2 〇 b回應於位 址位元A 0 / A 1。當位址位元A 0 / A 1從[〇,〇 ]增加至[1,1 ]時, 資料數元組M4/M5/M6/M7會依序變成可存取。然而,半導 體記憶裝置2 0 a不具有關於資料數元址μ 3之任何記憶單 元,並且沒有資料數元組從半導體記憶裝置2 〇 a中讀出。 假如資料數元組Μ 0 - Μ 3代表程控指令,則不正確的程控指 令碼會從受位址[A 1,A 0 ] = [ 1 , 1 ]指定之半導體記憶裝置 2 0 a之記憶單元中讀出。4 2 5 5 7 5 V. Description of the invention (7) Type of quantity. On the other hand, the two address bits A 0 / A 1 are supplied to the address pins a 0 / ai of the semiconductor memory device 20b, and the semiconductor memory device 2 01 provides information about four data bytes M 4 / Memory space of Μ 5 / Μ 6 / Μ 7 The data storage capacity of the semiconductor memory device 20b is expressed as. Therefore, the semiconductor memory device 20b is of a type having a data storage capacity of the n-th power. The semiconductor memory devices 20 a / 2 0 b each have a data output port 0 U T 1/0 U T 2 ′ and are connected to a common data bus B s!. The address decoder 2 0 c has an inverter 2 0 d. The address bit A 2 is supplied to the output enable / chip selection pin OE 1 of the semiconductor memory device 20 a as an output enable / chip selection signal. The inverter 2 0 d generates an inverted signal of the output enable / chip selection signal, and supplies the inverted signal to the output enable / chip selection pin 0E 2. Therefore, the address decoder 20c causes one of the semiconductor memory devices 20a / 20b to respond to the address bits A0 / A1. As shown in FIG. 6, the address bits A2, A1, and A0 are sequentially changed. When the address bit A 2 is at the logic “0” level, the address decoder 2 0 c will cause the semiconductor memory device 2 0 a to respond to the address bit A 0 / A 1. On the other hand, a logic 1 &quot; address bit A 2 allows another semiconductor memory device 2 0 b to respond to the address bits A 0 / A 1. When the address bits A 0 / A 1 increase from [〇, 〇] to [1,1], the data byte groups M4 / M5 / M6 / M7 will become accessible sequentially. However, the semiconductor memory device 20a does not have any memory cell with respect to the data element address µ3, and no data byte group is read out from the semiconductor memory device 20a. If the data tuples M 0-M 3 represent program-controlled instructions, the incorrect program-controlled instruction code will be changed from the memory unit of the semiconductor memory device 2 0 a specified by the address [A 1, A 0] = [1, 1] Read out.

第U頁 ’,4 2557 5 一_ 五、發明說明(8) 電腦程式人貝以電腦語言發展電腦程式’並且編澤器 使電腦程式轉換成程控指令碼之集合。基於在記憶系統中 位址係連續之假設下,程控指令碼寫入記憶系統之半導體 唯讀記憶裝置中。假使如圖5中所示受位址中之一指定之 記憶單元不存在,則程控指令碼會遺失,並且電腦程式會 變得不完整。 【發蚵概述】 ’ 因此之故,本發明之重要目的係提供位址空間管理 器,其指定部分位址空間至具有2的η次方以下之儲存容量 而無浪費的位址之資料儲存器。 本發明之另一重要目的係提供具有内建式位址空間管 理器之半導體記憶裝置。 本發明之又一重要目的係提供記憶系統,其連續地指 定一連串位址至一個以上之資料儲存器,該資料儲存器中 至少一個具有2的η次方以下之儲存容量。 依據本發明之一態樣,提供一種位址空間管理器,其 用以指定由(η 1 + η 2 )個位址'位元所定義之位址空間的位 址次空間至用以儲存in 1區段資料貢訊之資料儲存器’此處 nl 、η2與ml係第一自然數、第二自然數與第三自然數,並 且該ml係等於或者大於2(nl~l)且小於2nl,其包含位址轉換 器,用以轉換該η 1位址位元,並且供應經轉換之位址位元 至資料儲存器以使位址次空間定位於位址•空間内2η1_χ個位 置中之一上,此處X係滿足方程式2nl-ml = 2X,以及位址Page U ’, 4 2557 5 I_ V. Description of the invention (8) Computer programs: Human computers develop computer programs in computer languages’ and editors convert computer programs into a set of program-controlled instruction codes. Based on the assumption that addresses are continuous in the memory system, the program-controlled instruction code is written into the semiconductor read-only memory device of the memory system. If the memory unit designated by one of the addresses as shown in FIG. 5 does not exist, the program control script will be lost and the computer program will become incomplete. [Overview of development] 'For this reason, an important object of the present invention is to provide an address space manager, which specifies a portion of the address space to a data storage device with a storage capacity of 2 to the power of η without waste. . Another important object of the present invention is to provide a semiconductor memory device having a built-in address space manager. Another important object of the present invention is to provide a memory system which continuously assigns a series of addresses to more than one data storage, and at least one of the data storages has a storage capacity of 2 to the power of n or less. According to one aspect of the present invention, an address space manager is provided, which is used to specify an address subspace of an address space defined by (η 1 + η 2) address bits to store the in The data storage of the data of 1 segment Gongxun 'here nl, η2 and ml are the first natural number, the second natural number and the third natural number, and the ml is equal to or greater than 2 (nl ~ l) and less than 2nl , Which includes an address converter to convert the η 1 address bit, and supplies the converted address bit to a data storage so that the address subspace is positioned in one of the 2η1_χ positions in the address • space First, where X is the equation 2nl-ml = 2X, and the address

第〗2頁 425575&quot; 五、發明說明(9) 解碼器,用以從至少包含該η 2位址位元之位址信號產生解 碼信號,並且供應解碼信號至資料儲存器以便從儲存於位 址次空間中之該m 1區段資料資訊中選取次區段資料資訊。 依據本發明之另一態樣,提供一種半導體記憶裝置, 受供應以定義位址空間之(η 1 + η 2 )個位址位元,此處該 η 1與該η 2係第一自然數與第二自然數,其包含一個以上之 記憶庫,該記憶庫中之一儲存m 1區段資料資訊於位址空間 之位址次空間,此處該ΙΏ 1係第三自然數並且等於或者大於 2(n卜&quot;且小於2nI,位址轉換器轉換該η 1位址位元,並且供應 經轉換之位址位元至該一個以上記憶庫中之一,以使位址 次空間定位於在位址空間中個位置中之一,此處X滿足一 方程式2&quot;1-1111 = 2&quot;,以及位址解碼器,用以從包含該112位 址位元之位址信號產生選擇信號與解碼信號,以選擇信號 選擇地啟動一個以上之記憶庫,並且供應解碼信號至資料 儲存器以便從儲存於位址次空間中之該m 1區段資料資訊中 選取次區段資料資訊。 依據本發明之又一態樣,提供一種記憶系統,包含複 數個記憶單位,其分別由位址空間的位址次空間所指定而 用以儲存資料資訊,並且回應於位址信號而用以至少讀出 資料資訊,該複數個記憶單位中之一係儲存m 1區段資料資 訊,此處m 1係大於2的(η - 1 )次方且小於2的η次方並且η 係自然數,以及位址空間管理器,其回應於位址信號之較 高的位址位元而用以選擇性地賦能複數個記憶單位,選擇 性地轉換較高的位址位元以修正形成部份位址信號之位址Page 425425 &quot; V. Description of the invention (9) A decoder for generating a decoded signal from an address signal containing at least the n 2 address bits, and supplying the decoded signal to a data storage device for storage from the address Select the sub-section data information from the m 1 section data information in the sub-space. According to another aspect of the present invention, a semiconductor memory device is provided, which is provided with (η 1 + η 2) address bits defining an address space, where η 1 and η 2 are first natural numbers. And the second natural number, which includes more than one memory bank, one of which stores m 1 segment data information in the address space of the address space, where the Ώ 1 is a third natural number and is equal to or Greater than 2 (n &quot; and less than 2nI, the address converter converts the n 1 address bit and supplies the converted address bit to one of the more than one memory bank to locate the address in the subspace In one of the positions in the address space, here X satisfies a formula 2 &quot; 1-1111 = 2 &quot; and an address decoder for generating a selection signal from an address signal containing the 112 address bits And decode the signal, selectively activate more than one memory bank with the selection signal, and supply the decoded signal to the data storage to select the sub-segment data information from the m 1-segment data information stored in the address sub-space. Another aspect of the present invention provides a The memory system includes a plurality of memory units, which are respectively designated by the address subspace of the address space to store data information, and are used to read at least the data information in response to the address signal. The plurality of memory units One is to store the data information of the m 1 section, where m 1 is the power of (η-1) greater than 2 and the power of η less than 2 and η is a natural number, and the address space manager responds to bits. The higher address bits of the address signal are used to selectively enable multiple memory units, and selectively convert the higher address bits to modify the addresses forming part of the address signal

第13頁 五、發明說明(10) 位元而替代較高的位址位元,而以在複數個記憶單位中不 具有不可定址記憶單元之方式使位址次空間連續。 【圖示之簡單說明】 從下文中附有圖示之說明將可更加清楚地瞭解記憶系 統、半導體記憶裝置與位址轉換器之特徵與優點,其中之 圖示為: 圖1之區塊圖係顯示先前技術的記憶系統之配置; 圖2係顯示位址信號與受存取數元組間之關係; 圖3之電路圖係顯示第一先前技術之記憶系統之配 置; 圖4係顯示位址位元與受存取資料數元組間之關係; 圖5之電路圖係顯示第二先前技術之記憶系統之配 置; 圖6係顯示位址位元與受存取資料數元組間之關係; 圖7之電路圖係顯示依據本發明之記憶系統之配置; 圖8'係顯示由併入記憶系統中之位址轉換器所實現之 位址轉換; 圖9之電路圖係顯示依據本發明之另一記憶系統之配 置; 圖1 0係顯示位址指定至併入記憶系統中之半導體記憶 裝置上; 圖1 1之電路圖係顯示依據本發明之半導體記憶裝置之 電路組態;Page 13 V. Description of the invention (10) bits instead of higher address bits, and the address subspace is made continuous by not having unaddressable memory cells in the plurality of memory units. [Brief description of the icons] The features and advantages of the memory system, semiconductor memory device, and address converter can be more clearly understood from the description with the icons below. The icons are as follows: Block diagram of Figure 1 Figure 2 shows the configuration of the prior art memory system; Figure 2 shows the relationship between the address signal and the accessed byte; Figure 3 is a circuit diagram showing the configuration of the first prior art memory system; Figure 4 shows the address The relationship between bits and bytes of accessed data; Figure 5 is a circuit diagram showing the configuration of the second prior art memory system; Figure 6 is the relationship between address bits and bytes of accessed data; FIG. 7 is a circuit diagram showing a configuration of a memory system according to the present invention; FIG. 8 ′ is a diagram showing an address conversion implemented by an address converter incorporated in the memory system; FIG. 9 is a circuit diagram showing another embodiment according to the present invention Configuration of the memory system; FIG. 10 shows the address assigned to the semiconductor memory device incorporated in the memory system; FIG. 11 is a circuit diagram showing the circuit configuration of the semiconductor memory device according to the present invention;

第14頁 五、發明說明(11) 圖1 2 A至1 2 D係顯示導通之選用電晶體、外部位址位元 與内部位址位元; 圖丨3之電路圖係顯示依據本發明之另一半導體記憶裝 置之電路組態; 圖1 4 A與1 4 B係顯示控制信號、外部位址位元與内部位 址位元間之關係, 圖1 5 A與1 5 D係顯示控制信號與未指定入半導體記憶裝 置之位址次空間間之關係; 圖1 6之電路圖係顯示依據本發明之又一半導體記憶裝 置之電路組悲。 【符號之說明】 1〜記憶系統 2〜微處理器 1 0〜記憶系統 2 0〜記憶系統 3 0〜微處理器 31 ~ 記憶系統 3 1 a / 3 1 b〜半導體記憶裝置 3 1 c ~ 位址解碼器 3 1 d〜位址轉換器 3 1 e / 3 1 g〜記憶單元陣列 3 1 f / 3 1 h〜資料缓衝器 31j〜OR閘 3 1 k〜反相器Page 14 V. Description of the invention (11) Figures 1 2 to 1 2 D show the selection of transistors, external address bits and internal address bits that are turned on; the circuit diagram of Figure 3 shows another circuit according to the present invention. The circuit configuration of a semiconductor memory device; Figures 14A and 14B show the relationship between the control signals, the external address bits and the internal address bits, and Figures 15A and 15D show the control signals and The relationship between the sub-address space of the semiconductor memory device is not specified. The circuit diagram of FIG. 16 is a circuit diagram of another semiconductor memory device according to the present invention. [Explanation of symbols] 1 ~ memory system 2 ~ microprocessor 1 0 ~ memory system 2 0 ~ memory system 3 0 ~ microprocessor 31 ~ memory system 3 1 a / 3 1 b ~ semiconductor memory device 3 1 c ~ bit Address decoder 3 1 d ~ address converter 3 1 e / 3 1 g ~ memory cell array 3 1 f / 3 1 h ~ data buffer 31j ~ OR gate 3 1 k ~ inverter

第15頁 -425575^ 五、發明說明(12) 31m - 互斥-N 0 R閘 3 1 η〜反相器 3 2〜位址匯流排 3 3〜資料匯流排 4 1〜記憶系統 4 1 a / 4 1 b〜半導體記憶裝置 4 1 c ~ 位址解碼器 4 1 d ~ 位址轉換器 4 1 e / 4 1 g〜記憶單元陣列 4 1 f / 4 1 h〜資料缓衝器 41j/41k/41m 〜NOR 閘 4 1 η〜反相器 4 1 ρ - 反相器 41q ~ 互斥—NOR問 4 2〜位址匯流排 4 3〜資料匯流排 5 0〜半導體晶片 5 1 / 5 2 ~ 記憶庫 5 3〜位址解碼器 5 4〜可程式位址轉換器 54a/54b 〜NOR 閘 54c/54d ~ 反相器 54e〜互斥- NOR閘 54i/54g/54h〜反相器Page 15-425575 ^ V. Description of the invention (12) 31m-Mutual exclusion-N 0 R gate 3 1 η ~ inverter 3 2 ~ address bus 3 3 ~ data bus 4 1 ~ memory system 4 1 a / 4 1 b ~ semiconductor memory device 4 1 c ~ address decoder 4 1 d ~ address converter 4 1 e / 4 1 g ~ memory cell array 4 1 f / 4 1 h ~ data buffer 41j / 41k / 41m ~ NOR gate 4 1 η ~ Inverter 4 1 ρ-Inverter 41q ~ Mutual exclusion—NOR Question 4 2 ~ Address bus 4 3 ~ Data bus 5 0 ~ Semiconductor chip 5 1/5 2 ~ Memory bank 5 3 ~ address decoder 5 4 ~ programmable address converter 54a / 54b ~ NOR gate 54c / 54d ~ inverter 54e ~ mutually exclusive-NOR gate 54i / 54g / 54h ~ inverter

第16頁 425575 五、發明說明(13) 54j/54k/54m/54n/54p/54q 〜選用電晶體 54r/54s/54t/54u ~ 反相器 5 5〜資料緩衝器 5 6〜資料埠 6 0〜半導體晶片 6 1 / 6 2〜記憶庫 6 3〜位址解碼器 6 4〜位址轉換器 64a/64b/64c/64d ~ 反相器 64e/64f/64g/64h/64j/64k人64m/64n/64p/64q 〜NAND 閘 6 5〜資料緩衝器 7 0〜半導體晶片 7 1 / 7 2〜記憶庫 7 3 ~ 位址解碼器 - 7 4〜位址轉換器 7 5 ~ 資料緩衝器 【較佳實施例之詳細說明】 記憶系統 第一實施例 參照圖7,實施本發明之記憶系統3 1與微處理器3 0形 成資料處理系統之一部份,並且微處理器3 0係透過位址匯 流排3 2與資料匯流排3 3而聯絡於記憶系統3 1 。位址匯流排 32傳送23 -位元位址信號A00-A22至記憶系統31 ,而資料匯Page 16 425575 V. Description of the invention (13) 54j / 54k / 54m / 54n / 54p / 54q ~ Use transistor 54r / 54s / 54t / 54u ~ Inverter 5 5 ~ Data buffer 5 6 ~ Data port 6 0 ~ Semiconductor wafer 6 1/6 2 ~ Memory bank 6 3 ~ Address decoder 6 4 ~ Address converter 64a / 64b / 64c / 64d ~ Inverter 64e / 64f / 64g / 64h / 64j / 64k people 64m / 64n / 64p / 64q ~ NAND gate 6 5 ~ data buffer 7 0 ~ semiconductor chip 7 1/7 2 ~ memory bank 7 3 ~ address decoder-7 4 ~ address converter 7 5 ~ data buffer [Compared Detailed description of the preferred embodiment] Referring to FIG. 7, the first embodiment of the memory system, the memory system 31 and the microprocessor 30 implementing the present invention form a part of a data processing system, and the microprocessor 30 is through an address The bus 32 2 and the data bus 33 are connected to the memory system 3 1. The address bus 32 sends a 23-bit address signal A00-A22 to the memory system 31, and the data sink

-〇〇 /5&quot;· 五、發明說明(14) 流排3 3從記憶系統3 1傳送1 6 -位元資料碼至微處理器3 0。 在下文之說明中’名詞「數元組」意指1 6 -位元資料碼。 記憶系統3 1包含半導體記憶裝置3 1 a / 3 1 b、位址解碼 器3 1 c與位址轉換器3 1 d。位址位元A 2 1 / A 2 2係供應至位址 解碼器3 1 c,並且位址解碼器3 1 c係藉著低態有效位準之外 部晶片選擇/輸出賦能信號而選擇性地賦能半導體記憶裝 置31a/31b。位址轉換器3ld經轉換之位址位元A21/A22至 位址位元Α2Γ/Α22 ’並且供應位址位元Α2Γ/Α22,至半導 體記憶裝置3 1 b。因此’記憶系統3丨之位址空間係區分為 二個位址次空間’,且半導體記憶裝置3丨a/ 3丨b係分別地 提供由較低位址所疋義之位址次空間以及由較高位址所定 義之位址次空間。 當位址解碼器3 1 c藉著低態有效位準之外部晶片選擇/ 輸出賦能信號而賦能半導體記憶裝置3丨a時,半導體記憶 裝置31a會回應於位址位元aqo — A2〇,並且傳送資料碼至資 料匯流排3 3。另一方面’當位址解碼器3丨^藉著低態有效 位準之外部晶月選擇/輸出賦能信號而賦能另—半導體記 憶裝置3 1 b時’半導體記憶裝置3丨b會回應於位址位元 A 0 0 - A 2 0 / A 2 1 / A 2 2 ,並且傳送資料碼至資料匯流排3 3。 半導體記憶裝置3 1 a包含記憶單元陣列3.1 e與資料缓衝 器31 f,並且位址埠API 、控制信號埠〇E】與資料埠⑽了丨皆 併入半導體記憶裝置3 1 a中》記憶單元陣列3丨e包含2百萬 =1 6位兀個可定址記憶單元,並且提供關於2百萬數元組 資料或32百萬位元之記憶次空間。半導體記憶裝置3丨a之-〇〇 / 5 &quot; V. Description of the invention (14) The stream line 3 3 transmits a 16-bit data code from the memory system 3 1 to the microprocessor 30. In the following description, the term "math" means a 16-bit data code. The memory system 3 1 includes a semiconductor memory device 3 1 a / 3 1 b, an address decoder 3 1 c, and an address converter 3 1 d. The address bits A 2 1 / A 2 2 are supplied to the address decoder 3 1 c, and the address decoder 3 1 c is selectively selected by an external chip select / output enable signal with a low active level Ground-enabling semiconductor memory devices 31a / 31b. The address converter 3ld converts the address bits A21 / A22 to the address bits A2Γ / Α22 'and supplies the address bits A2Γ / Α22 to the semiconductor memory device 3 1b. Therefore, 'the address space of the memory system 3 丨 is divided into two address subspaces', and the semiconductor memory device 3 丨 a / 3 丨 b respectively provide the address subspace defined by the lower address and Address subspace defined by higher addresses. When the address decoder 3 1 c enables the semiconductor memory device 3 丨 a by an external chip selection / output enable signal with a low active level, the semiconductor memory device 31 a responds to the address bits aqo — A2. And send a data code to the data bus 3 3. On the other hand, 'when the address decoder 3 丨 ^ is enabled by an external crystal moon selection / output enable signal with a low state effective level, the other semiconductor memory device 3 1b will respond. At the address bits A 0 0-A 2 0 / A 2 1 / A 2 2, and send a data code to the data bus 33. The semiconductor memory device 3 1 a includes a memory cell array 3.1 e and a data buffer 31 f, and the address port API, the control signal port 0E] and the data port are all incorporated into the semiconductor memory device 3 1 a. The cell array 3e includes 2 million = 1 6-bit addressable memory cells, and provides a memory subspace of 2 million bytes of data or 32 million bits. Semiconductor memory device 3 丨 a

第18頁 _ 4 -25 5.7LS- 五、發明說明(15) 記憶次空間係表示為「2的η次方」。資料緩衝器3 1 f係連 接於記憶單元陣列3 1 e與資料埠0 U T 1之間,並且資料埠 ◦ U T 1係連接至資料匯流排3 3。位址位元A 0 0 - A 2 0係供應於 位址埠AP 1 ,並且外部晶片選擇/輸出賦能信號係供應至控 制信號埠〇E 1。 當控制信號埠Ο E 1處之外部晶片選擇/輸出賦能信號改 變成低態有效位準時,1 6 -位元資料碼會從受位址指定之 記憶單元中讀出,該位址係等同於由位址位元A 0 0 - A 2 0所 表示之位址,並且轉移至資料緩衝器3 1 f。因此,1 6 -位元 資料碼會從資料缓衝器3 1 ί傳送至資料匯流排3 3。 另一半導體記憶裝置3 1 b包含記憶單元陣列3 1 g與資料γ 緩衝器31h,並且位址埠ΑΡ2、控制信號埠0Ε2與資料埠 0UT2皆併入半導體記憶裝置31b中。記憶單元陣列3 lg包含 6百萬X 1 6位元個可定址記憶單元,並且提供關於6百萬數 元組資料或9 6百萬位元之記憶次空間。半導體記憶裝置 3 1 b之記憶次空間係大於2的(η- 1 )次方且小於2的η次方。 因此之故,記憶單元陣列3 1 g提供2的η次方以下之記憶次 空間。資料緩衝器3 1 h係連接於記憶單元陣列3 i g與資料埠 0UT2之間,並且資料埠0UT2係連接至資料匯流排33。位址 位元八00-八20/八21’/八22’係供應於位址埠八?2,並且外部晶 片選擇/輸出賦能信號係供應至控制信號埠0E2。 當控制信號埠0E2處之外部晶片選擇/輸出賦能信號改 變成低態有效位準時,1 6 -位元資料碼會從受位址指定之 記憶單元中讀出,該位址係等同於由位址位元A 0 0 - A 2 0所Page 18 _ 4 -25 5.7LS- V. Description of the Invention (15) The memory subspace is expressed as "2th power of η". The data buffer 3 1 f is connected between the memory cell array 3 1 e and the data port 0 U T 1, and the data port ◦ U T 1 is connected to the data bus 33. The address bits A 0 0-A 2 0 are supplied to the address port AP 1, and the external chip select / output enable signal is supplied to the control signal port 0E 1. When the external chip select / output enable signal at the control signal port 0 E 1 is changed to the low effective level, the 16-bit data code will be read from the memory unit designated by the address, which is equivalent to At the address represented by address bits A 0 0-A 2 0, and transfer to the data buffer 3 1 f. Therefore, a 16-bit data code is transmitted from the data buffer 3 1 to the data bus 3 3. The other semiconductor memory device 31b includes a memory cell array 31g and a data gamma buffer 31h, and the address port AP2, the control signal port 0E2, and the data port OUT2 are all incorporated into the semiconductor memory device 31b. The memory cell array 3 lg contains 6 million X 1 6-bit addressable memory cells and provides about 6 million bytes of data or 9 6 million bits of memory subspace. The memory subspace of the semiconductor memory device 3 1 b is greater than the (η− 1) th power and less than the nth power. For this reason, the memory cell array 31 g provides a memory subspace of 2 to the power of n. The data buffer 31 h is connected between the memory cell array 3 i g and the data port OUT2, and the data port OUT2 is connected to the data bus 33. Address Bits 00-80 20 / Eight 21 '/ Eight 22' are supplied to the address port eight? 2, and the external chip select / output enable signal is supplied to the control signal port 0E2. When the external chip select / output enable signal at the control signal port 0E2 is changed to the low effective level, the 16-bit data code will be read from the memory unit designated by the address, which is equivalent to Address Bits A 0 0-A 2 0

第19頁 v 425575&quot; —_____ , ' &quot; 1 I - . - _ _ 五、發明說明(16) &quot; -- ί不之位址’並且轉移至資料緩衝器3 1 h。因此,1 6 -位元 貢料碼會從資料緩衝器3 1 h傳送至資料匯流排33。 a 位址解碼器3 1 c包含〇R閘3 1 j與反相器3 1 k,並且外部 晶片選擇/輸出賦能信號係從OR閘31 j與反相器31k選擇地 供應至控制信號埠0 E 1與另一控制信號埠〇 E 2。位址位元 A 21/ A 2 2係供應至〇 R閘3 I j之輸入節點,並且〇 R閘3 1 j之輸 出節點係連接至控制信號埠〇E 1與反相器3 1 k之輸入節點。 反相器3 1 k之輸出節點係連接至控制信號埠〇 £ 2。 當位址位元A 2 1 / A 2 2二者皆為低位準時,〇R閘3 1 j會供 應低態有效位準之外部晶片選擇/輸出賦能信號至控制信 號埠0E1 ’並且使半導體記憶裝置3ia回應於位址位元 AOO-A20。然而,另一半導體記憶裝置3U並未回應於位址 信號。因此’半導體記憶裝置3丨a提供由外部位址信號[〇, 0 ’ *,*,….·,* ]所代表之記憶次空間。星號[* ]代表&quot;丨”或 丨丨〇&quot;。 另一方面,假使位址位元A 2 1 / A 2 2中至少一個係在高 位準’則0 R閘3 1 j會供應高態無效位準之外部晶片選擇/輸 出賦能信號至控制信號埠〇 E 1 ’並且反相器3 1 k會供應高態 有效位準之外部晶片選擇/輸出賦能信號至控制信號埠 0 E 2。半導體記憶裝置3 1 a並未回應於位址位元A 〇 〇 - A 2 0, 而另一半導體記憶裝置31b從記憶單元傳送16-位元資料碼 至資料匯流排3 3,該記憶单元係由位址位元 Α00-Α20/Α2Γ/Α22’所特定。因此,半導體記憶裝置3lb提 供由位址信號[1,1,*,*,.... .,* ]、[ 〇, 1,*,*,…_ ·,* ]或Page 19 v 425575 &quot; —_____, '&quot; 1 I-.-_ _ V. Description of the Invention (16) &quot;-ί 不 之 Address' and transfer to the data buffer 3 1 h. Therefore, the 16-bit code is sent from the data buffer 31 h to the data bus 33. a The address decoder 3 1 c includes an OR gate 3 1 j and an inverter 3 1 k, and the external chip select / output enable signal is selectively supplied to the control signal port from the OR gate 31 j and the inverter 31k. 0 E 1 and another control signal port 0E 2. The address bits A 21 / A 2 2 are supplied to the input node of the OR gate 3 I j, and the output node of the OR gate 3 1 j is connected to the control signal port OE 1 and the inverter 3 1 k. Enter the node. The output node of the inverter 31k is connected to the control signal port 0 £ 2. When the address bits A 2 1 / A 2 2 are both low, the OR gate 3 1 j will supply an external chip select / output enable signal with a low effective level to the control signal port 0E1 'and enable the semiconductor The memory device 3ia responds to the address bits AOO-A20. However, another semiconductor memory device 3U did not respond to the address signal. Therefore, the 'semiconductor memory device 3a' provides a memory subspace represented by the external address signals [0, 0 '*, *, ...., *]. The asterisk [*] stands for "quote" or "丨 丨". On the other hand, if at least one of the address bits A 2 1 / A 2 2 is at a high level, then 0 R gate 3 1 j will supply high The external chip select / output enable signal of the inactive level is controlled to the control signal port 0E 1 'and the inverter 3 1 k will supply the external chip select / output enable signal of the high state active level to the control signal port 0 E 2. The semiconductor memory device 3 1 a does not respond to the address bits A 〇〇- A 2 0, and the other semiconductor memory device 31b transmits a 16-bit data code from the memory unit to the data bus 33, the memory The unit is specified by the address bits A00-Α20 / Α2Γ / Α22 '. Therefore, the semiconductor memory device 3lb provides the address signals [1, 1, *, *, ...,..., *], [〇, 1, *, *, ..._ ·, *] or

第20頁 425575 五、發明說明(17) [1,0,*,*,…..,* ]所代表之記憶次空間。 位址轉換器3 1 d包含互斥-NOR閘3 1 m與反相器3 1 η。位 址位元A 2 1 / A 2 2係供應至互斥-NOR閘3 1 m之輸入節點,並且 位址位元A 2 1係供應至反相器3 1 η之輸入節點。互斥-N 0 R閘 3 1 m與反相器3 1 η係供應位址位元A 2 2 ’與A 2 1 ’至半導體記憶, 裝置31b之位址埠AP2。 雖然互斥-N 0 R閘3 1 m從皆為邏輯π 0 &quot;位準之位址位元 Α22、Α21產生邏輯” 〇&quot;位準,位址位元Α22、Α21之組合使 外部晶片選擇/輸出賦能信號去能半導體記憶裝置3 1 b,並 且不考慮位址位元[Α22,Α21] = [〇,〇]之組合。因此,互斥 -NOR閘31m從位址位元Α21/Α22產生位址位元Α2Γ/Α22’之 γ 三種組合,並且如圖8所示位址轉換器3 1 d變化位址位元 Α22’/Α2Γ 。介於位址位元A22/A21與位址位元Α2Γ/Α22’ 間之邏輯關係表示為 Α22, =Α22 ·Α21 + A 22 A 21 ,A2Γ 二 A 21 其中A意指A之反相,通常係以在A上方畫線來表示。位址 位元A22’ / A21’之組合係代表等於[A22, A21 ] - 1之位址。 皆為邏輯” 〇&quot;位準之位址位元A22/A21係代表半導體記 憶裝置31a,並且位址位元A22/A21之組合不作為記憶單元 陣列3 1 g上之位址。倘若位址轉換器3 1 d並未併入記憶系統 3 1 ,則類似圖3所示之先前技術,電腦設計者便不能使用 半導體記憶裝置31b中以[A22, A21,A20,…· ,A00]= [〇,〇,*…·. * ]表示之位址所指定之記憶單元。 本發明人注意到因為記憶次空間係9 6百萬-位元,所Page 20 425575 V. Description of the invention (17) The memory subspace represented by [1, 0, *, *, ....., *]. The address converter 3 1 d includes a mutex-NOR gate 3 1 m and an inverter 3 1 n. The address bit A 2 1 / A 2 2 is supplied to the input node of the mutex-NOR gate 3 1 m, and the address bit A 2 1 is supplied to the input node of the inverter 3 1 η. The mutually exclusive-N 0 R gate 3 1 m and the inverter 3 1 η supply address bits A 2 2 ′ and A 2 1 ′ to the semiconductor memory, and the address port AP2 of the device 31 b. Although the mutually exclusive -N 0 R gate 3 1 m is logical from all the π 0 &quot; level address bits A22, A21 generates logic "〇 &quot; level, the combination of address bits A22, A21 makes the external chip The selection / output enable signal disables the semiconductor memory device 3 1 b, and does not consider the combination of address bits [Α22, A21] = [〇, 〇]. Therefore, the mutually exclusive-NOR gate 31m is derived from the address bit A21 / Α22 generates three combinations of address bits A2Γ / Α22 ', and the address converter 3 1 d changes the address bits A22' / Α2Γ as shown in Fig. 8. Between the address bits A22 / A21 and the bit The logical relationship between the address bits A2Γ / Α22 'is expressed as A22, = Α22 · A21 + A 22 A 21, A2Γ and A 21 where A means the inverse of A, which is usually expressed by drawing a line above A. Bit The combination of address bits A22 '/ A21' represents an address equal to [A22, A21]-1. All are logical "〇 &quot; The address bits A22 / A21 represent the semiconductor memory device 31a, and the bit The combination of address bits A22 / A21 is not used as an address on the memory cell array 3 1 g. If the address converter 3 1 d is not integrated into the memory system 3 1, similar to the prior art shown in FIG. 3, the computer designer cannot use the semiconductor memory device 31 b with [A22, A21, A20, ..., A00 ] = [〇, 〇, * ... ·. *] Indicates the memory unit specified by the address. The inventor noticed that because the memory subspace is 96 million-bits,

第21頁Page 21

/Μ 五、發明說明(18) 以半導體記憶裝置3 1 b並未具有任何表示為 [A22, A21,A20 …. A00] = [1,1,* …· *]之記憶單元。為 了 解救以[A22,A21,A20,…..,a〇〇] = [〇,〇,* …·,*]表不 之位址所指定之記憶單元’本發明人提供位址轉換器 31d,並且使位址轉換器31d依序將表示為[A22, A21 ]= [1,1]、[Α22,Α21] = [1,0]與[A22,A21] = [0,1]之位址 轉換成表示為[A22,A21] = [1,0]、[A22,A21] = [0,1]與 [A 2 2 , A 2 1 ] = [ 0,0 ]之位址。因此,未指定之位址,亦即 [A 2 2,A 2 1 ] = [ 1,1 ]會轉換成業已指定至記憶單元之位 址,亦即[A22,A21] = [1,〇],並且位址[A22,A21] = [0,1] 會指定至未被位址[A 2 2 , A 2 1 ] = [ 0,0 ]所特定之記憶單 元。 從前述之發明中將瞭解,依據本發明之位址轉換器 3 1 d係依序地移位由外部位址所表示之位址,並且使半導 體記憶裝置3 1 b、之所有記憶單元皆可利用D微處理器3 〇可 輕易地增加位址。 半導體記憶裝置3 1 b得具有缺陷記憶單元,據此,僅 有9 6百萬-位元記憶單元存留於記憶單元陣列3 1 g中。然 而,位址轉換器3 1 d解救此等部分缺陷的半導體記憶裝置 3 1 b,並且資料處理系統之生產成本會減低。 在此例子中,位址解碼器3 1 c與位址轉換器3 1 d整體構 成位址空間管理器。位址轉換器3〗d之函數廣義而言即: 由位址位元A 2 1 / A 2 2所代表之值減去在記憶單元陣列3丨e中 最高位址所代表之值。 »/ M 5. Description of the invention (18) The semiconductor memory device 3 1 b does not have any memory unit represented as [A22, A21, A20…. A00] = [1, 1, *… · *]. In order to understand how to save the memory unit specified by [A22, A21, A20,… .., a〇〇] = [〇, 〇, *… ·, *], the present inventor provides an address converter 31d , And the address converter 31d will be sequentially expressed as [A22, A21] = [1,1], [Α22, A21] = [1,0], and [A22, A21] = [0,1] The addresses are converted into addresses expressed as [A22, A21] = [1, 0], [A22, A21] = [0, 1], and [A 2 2, A 2 1] = [0, 0]. Therefore, an unspecified address, that is, [A 2 2, A 2 1] = [1,1] will be converted into an address that has been assigned to the memory unit, that is, [A22, A21] = [1, 0] , And the address [A22, A21] = [0,1] will be assigned to the memory unit not specified by the address [A 2 2, A 2 1] = [0,0]. It will be understood from the foregoing invention that the address converter 3 1 d according to the present invention sequentially shifts the address indicated by the external address, and makes all the memory cells of the semiconductor memory device 3 1 b. Addresses can be easily added using the D microprocessor 30. The semiconductor memory device 31b has a defective memory cell, and accordingly, only 96 million-bit memory cells remain in the memory cell array 31g. However, the address converter 3 1 d rescues these partially defective semiconductor memory devices 3 1 b, and the production cost of the data processing system is reduced. In this example, the address decoder 3 1 c and the address converter 3 1 d integrally constitute an address space manager. The function of the address converter 3 d is, in a broad sense, the value represented by the address bits A 2 1 / A 2 2 minus the value represented by the highest address in the memory cell array 3 丨 e. »

第22頁 425575 五、發明說明(19) 第二實施例 參照圖9 ,實施本發明之另一記憶系統4丨包含半導體 記憶裝置41a/41b、位址解碼器41c與位址轉換器41d。記 憶系統4 1與微處理器(未顯示出)形成資料處理系統之一 部伤,並且微處理器係透過位址匯流排4 2與資料匯流排4 3 而聯絡於記憶系統4 1 。位址匯流排42傳送24_位元位址信 號A 0 0 - A 2 3至記憶系統4 1 ,而資料匯流排4 3從記憶系統4】 傳送位元資料碼至微處理器。在下文之說明中,名詞 「數元組」亦意指1 6 -位元資料碼。 記憶系統4 1包含半導體記憶裝置4丨a / 4丨b、位址解碼 器4 1 c與位址轉換器4 1 d。位址位元A 2丨/ A 2 2 / A 2 3係供應至 位址解碼器4 1 c ’並且位址解碼器4丨c係藉著低態有效位準 之外部晶片選擇/輸出賦能信號而選擇性地賦能半導體記 憶裝置4 1 a / 4 1 b。位址轉換器4 1 d經轉換之位址位元 A21/A22至位址位元Α2Γ/Α22,。位址位元A00-A22係供應 至半導體記憶裝置41a ’並且位址位元α〇〇-Α22/Α2Γ/Α22, 係從位址匯流排4 2與位址轉換器4 1 d供應至半導體記憶裝 置4丨b。因此’記憶系統4 1之位址空間係區分為二個位址 次空間。 當位址解碼器4 1 c藉著低態有效位準之外部晶片選擇/ 輸出賦能信號而賦能半導體記憶裝置4丨a時’半導體記憶 裝置41a會回應於位址位元A〇〇_A22,並且傳送〗6_位元資 料,至資料匯流排43。另一方面,當位址解碼器4〗c藉著 低態有效位準之外部晶片選擇/輸出賦能信號而賦能另一Page 22 425575 V. Description of the Invention (19) Second Embodiment Referring to FIG. 9, another memory system 4 implementing the present invention includes a semiconductor memory device 41a / 41b, an address decoder 41c, and an address converter 41d. The memory system 41 and the microprocessor (not shown) form part of the data processing system, and the microprocessor is connected to the memory system 41 through the address bus 4 2 and the data bus 4 3. The address bus 42 transmits a 24-bit address signal A 0 0-A 2 3 to the memory system 41, and the data bus 43 transmits the bit data code from the memory system 4 to the microprocessor. In the following description, the term "matrix" also means a 16-bit data code. The memory system 41 includes a semiconductor memory device 4a / 4b, an address decoder 41c, and an address converter 41d. The address bits A 2 丨 / A 2 2 / A 2 3 are supplied to the address decoder 4 1 c 'and the address decoder 4 丨 c is enabled by an external chip selection / output enable at a low active level Signals to selectively energize the semiconductor memory device 4 1 a / 4 1 b. The address converter 4 1 d converts the address bits A21 / A22 to the address bits A2Γ / Α22 ,. The address bits A00-A22 are supplied to the semiconductor memory device 41a 'and the address bits α〇〇-Α22 / Α2Γ / Α22 are supplied from the address bus 4 2 and the address converter 4 1 d to the semiconductor memory Device 4 丨 b. Therefore, the address space of the 'memory system 41' is divided into two address subspaces. When the address decoder 4 1 c enables the semiconductor memory device 4 a through the external chip selection / output enable signal of the low-active level, the semiconductor memory device 41 a will respond to the address bit A〇〇_ A22, and transmit 6_bit data to data bus 43. On the other hand, when the address decoder 4 is enabled by another external chip select / output enable signal with low active level

第23頁 ;4 2 5 5 7 5 &quot;_^_ 五'發明說明(20) 半導體記憶裝置4 1 b時,半導體記憶裝置4 1 b會回應於位址 位元A 0 0 - A 2 0 / A 2 1 ’/ A 2 2 ’,並且傳送1 6 -位元資料碼至資料 匯流排4 3 。 半導體記憶裝置4 1 a包含記憶單元陣列4 1 e與資料緩 衝器41 f,並且位址埠AP3、控制信號埠OE3與資料埠OUT3 皆併入半導體記憶裝置4 1 a中。記憶單元陣列4 1 e包含6百 萬X 1 6位元個可定址記憶單元,並且提供關於6百萬數元 組資料或9 6百萬位元之記憶次空間。記憶單元陣列4 1 e係 落於2的(η - 1 )次方與2的η次方間之範圍内,並且半導體 記憶裝置4 1 a之記憶次空間係一種表示為2的η次方以下之 資料儲存器。資料緩衝器4 1 ί係連接於記憶單元陣列4 1 e與 資料埠OUT3之間,並且資料埠OUT3係連接至資料匯流排 43。位址位元A00-A22係供應於位址埠AP3,並且外部晶片 選擇/輸出賦能信號係供應至控制信號埠OE 3。 當控制信號埠OE3處之外部晶片選擇/輸出賦能信號改 變成低態有效位準時,1 6 -位元資料碼會從受位址指定之 記憶單元中讀出,該位址係等同於由位址位元A00-A2 2所 表示之位址,並且轉移至資料缓衝器4 1 f。從外部晶片選 擇/輸出賦能信號產生内部晶片選擇/輸出賦能信號IOE3 , 並且業已賦能資料缓衝器4 1 f。因此,1 6 -位元資料碼係從 資料緩衝器4 1 ί傳送至資料匯流排4 3。 另一半導體記憶裝置4 1 b包含記憶單元陣列4 1 g與資料 缓衝器41h,並且位址埠AP4、控制信號埠0E4與資料埠 OUT4皆併,入半導體記憶裝置41b中。記憶單元陣列41g包含Page 23; 4 2 5 5 7 5 &quot; _ ^ _ Five 'invention description (20) When the semiconductor memory device 4 1 b, the semiconductor memory device 4 1 b will respond to the address bits A 0 0-A 2 0 / A 2 1 '/ A 2 2', and send a 16-bit data code to the data bus 4 3. The semiconductor memory device 41a includes a memory cell array 41e and a data buffer 41f, and the address port AP3, the control signal port OE3, and the data port OUT3 are all incorporated into the semiconductor memory device 41a. The memory cell array 4 1 e contains 6 million X 1 6-bit addressable memory cells and provides about 6 million bytes of data or 9 6 million bits of memory subspace. The memory cell array 4 1 e falls within a range between the (η-1) power of 2 and the η power of 2 and the memory subspace of the semiconductor memory device 4 1 a is a type of η or less expressed as 2 Data storage. The data buffer 41 is connected between the memory cell array 41 and the data port OUT3, and the data port OUT3 is connected to the data bus 43. Address bits A00-A22 are supplied to address port AP3, and an external chip select / output enable signal is supplied to control signal port OE3. When the external chip select / output enable signal at the control signal port OE3 is changed to the low effective level, the 16-bit data code will be read from the memory unit designated by the address, which is equivalent to The addresses represented by bits A00-A2 2 are transferred to the data buffer 4 1 f. The internal chip select / output enable signal IOE3 is generated from the external chip select / output enable signal, and the data buffer 41f has been enabled. Therefore, the 16-bit data code is transmitted from the data buffer 41 to the data bus 43. The other semiconductor memory device 41b includes a memory cell array 41g and a data buffer 41h, and the address port AP4, the control signal port 0E4, and the data port OUT4 are all merged into the semiconductor memory device 41b. Memory cell array 41g contains

第24頁 五、發明說明(21) 6百萬X 1 6位元個可定址記憶單元,並且提供關於6百萬數 元組資料或9 6百萬位元之記憶次空間。半導體記憶裝置 4 1 b之記憶次空間係大於2的(η - 1 )次方且小於2的η次 方。因此之故,記憶單元陣列4 1 g係一種在2的η次方以下 之資料儲存器。資料缓衝器4 1 h係連接於記憶單元陣列4 1 g 與資料埠0UT4之間,並且資料埠0UT4係連接至資料匯流排 43。位址位元Α00-Α20/Α2Γ/Α22’係供應於位址皡AP4,並 且外部晶片選擇/輸出賦能信號係供應至控制信號埠〇E4。 當控制信號埠0E4處之外部晶片選擇/輸出賦能信號改 .變成低態有效位準時,1 6 -位元資料碼會從受位址指定之 記憶單元中讀出,該位址係.等同於由位址位元 Α00-Α20/Α21’/Α22’所表示之位址,並且轉移至資料緩衝 ii41h。從外部晶片選擇/輸出賦能信號產生内部晶片選擇 /輸出賦能信號I Ο E 4 ’並且業已賦能資料緩衝器4 1 h。因 此,1 6 -位元資料碼係從資料緩衝器4 1 h傳送至資料匯流排 43 ° 位址解碼器4 1 c包含N 0 R閘4 1 j / 4 1 k / 4 1 m與反相器4 1 η, 並且外部晶片選擇/輸出賦能信號係從Ν 〇 R閘4 1 m與反相器 41η選擇地供應至控制信號埠OE3與另一控制信號埠0Ε4 Ί 位址位元A21/A22/A23係選擇地供應至n〇R閘41j/41k之輸 入節點’並且N 0 R閘4 1 j / 4 1 k之輸出節點係連接至另一 N 〇 r 閘4 1 m之輸入.郎點。N 0 R閘4 1 m之輸出節點係連接至控制信 號埠0E3與反相器4 I η之輸入節點。反相器4 1 n之輸出節點 係連接至控制信號埠0E4。Page 24 V. Description of the invention (21) 6 million X 1 6-bit addressable memory units, and provide about 6 million bytes of data or 9 6 million bits of memory subspace. The memory subspace of the semiconductor memory device 4 1 b is greater than 2 (η-1) power and less than 2 η power. For this reason, the memory cell array 41g is a kind of data storage device which is less than or equal to the power of n. The data buffer 41 h is connected between the memory cell array 41 g and the data port OUT4, and the data port OUT4 is connected to the data bus 43. The address bits A00-Α20 / Α2Γ / Α22 ′ are supplied to the address 皡 AP4, and the external chip select / output enable signal is supplied to the control signal port 0E4. When the external chip selection / output enable signal at the control signal port 0E4 is changed. When it becomes the low effective level, the 16-bit data code will be read from the memory unit designated by the address, which is equivalent to At the address represented by address bits A00-Α20 / Α21 '/ Α22', and transferred to the data buffer ii41h. The internal chip select / output enable signal I 0 E 4 ′ is generated from the external chip select / output enable signal and the data buffer 41 h has been enabled. Therefore, the 16-bit data code is transmitted from the data buffer 41 h to the data bus 43 ° The address decoder 4 1 c contains the N 0 R gate 4 1 j / 4 1 k / 4 1 m and the inverse And the external chip select / output enable signal is selectively supplied from the NOR gate 4 1 m and the inverter 41 η to the control signal port OE3 and another control signal port 0E4 Ί address bit A21 / A22 / A23 is selectively supplied to the input node of gate 41j / 41k 'and the output node of N 0 R gate 4 1 j / 4 1 k is connected to the input of another gate 4 1 m. point. The output node of N 0 R gate 4 1 m is connected to the input node of control signal port 0E3 and inverter 4 I η. The output node of the inverter 4 1 n is connected to the control signal port 0E4.

425575^ __ 五、發明說明¢22) 當位址位元A 2 1 / A 2 3二者皆係低位準時,N 〇 R閘4丨】會 改變其輸出節點為高位準,並且在輸入節點之一或二者出 現高位準下保持輸出節點為低位準。同樣地,當位址位元 A22/A23二者皆係低位準時,NOR閘41k會改變其胃輸出節點 為高位準’並且位址位元A22/A23之其他紐合會使N〇R問 41k產生低位準。當NOR閘41 j/41k改變其輸出節點為/位 準時’ NOR閘4 1 m會改變其輸出節點為高位準,並且反相器 4 1 m會供應低態有效位準之外部晶片選擇/輸出賦能化號^ 控制信號埠0E4。在NOR閘4 1j/41k之輸出節點處之 準之其他組合會使Ν 0 R閘4 1 m改變其輸出節點為低位準,並 且低態有效位準之外部晶片選擇/輸出賦能信號從N〇R閑' 4 1 m供應至信號控制埠0E 3。因此,位址解碼器4丨c係依據 位址位元A23/A22/A21而選擇地啟動半導體記,陳裝置 4 1 a / 4 1 b。 介於位址位元A23/A22/A21與受啟動之半導體記憶裝 置4 1 a/4 1 b間之關係係顯示於圓1 0中。假如位址位元 A 2 3 / A 2 2 / A 2 1係落於[0,0,0 ]與[0,1,0 ]間之範圍内,則位 址解碼器4 1 c會啟動半導體記憶裝置4 1 a。另一方面,假如 位址位元A23/A22/A21係落於[0, 1,1]與[1,0,1]間之範圍 内,則位址解碼器4 1 c會供應低態有效位準之外部晶片選 擇/輸出賦能信號至半導體記憶裝置4 1 b。 位址轉換器41d包含互斥-NOR閘41q與反相器41p。位 址位元A21/A22係供應至互斥-NOR閘41q之輸入節點’並且 位址位元A21係供應至反相器41p之輸入節點。互斥-NOR閘425575 ^ __ V. Description of the invention ¢ 22) When the address bits A 2 1 / A 2 3 are both low level, N 〇 gate 4 丨] will change its output node to a high level, and at the input node Keep the output node low when one or both levels are high. Similarly, when both address bits A22 / A23 are at low level, NOR gate 41k will change its stomach output node to high level 'and other kinks of address bit A22 / A23 will make NOR ask 41k Generate low level. When the NOR gate 41 j / 41k changes its output node to / level ', the NOR gate 4 1 m will change its output node to a high level, and the inverter 4 1 m will supply the external chip selection / output of the low state effective level Enabling number ^ control signal port 0E4. Other combinations of the standard at the output node of the NOR gate 4 1j / 41k will cause the N 0 R gate 4 1 m to change its output node to a low level, and the external chip select / output enable signal of the low-state effective level from N 〇R idle '4 1 m is supplied to the signal control port 0E 3. Therefore, the address decoder 4 丨 c selectively starts the semiconductor memory according to the address bits A23 / A22 / A21, and the device 4 1 a / 4 1 b. The relationship between the address bits A23 / A22 / A21 and the activated semiconductor memory device 4 1 a / 4 1 b is shown in circle 10. If the address bits A 2 3 / A 2 2 / A 2 1 fall within the range between [0,0,0] and [0,1,0], the address decoder 4 1c will start the semiconductor Memory device 4 1 a. On the other hand, if the address bits A23 / A22 / A21 fall within the range between [0, 1,1] and [1, 0, 1], the address decoder 4 1 c will supply the low state valid A level external chip select / output enable signal to the semiconductor memory device 4 1 b. The address converter 41d includes a mutex-NOR gate 41q and an inverter 41p. The address bit A21 / A22 is supplied to the input node of the mutex-NOR gate 41q 'and the address bit A21 is supplied to the input node of the inverter 41p. Mutual exclusion-NOR gate

第26頁 425575 五 '發明說明(23) 4 lq與反相器4 lp係供應位址位元A 22’與Α2Γ至半導體記憶 裝置41b之位址埠AP4。位址轉換器41 d之函數係表示為 A22’ = A22 · A 21 + A 22 · A21, A21’ = A 21 因此,位址轉換器4 1 d係將位址位元[A 2 2,A 2 1 ] = [ 1,1 ]、 [〇,〇 ]與[Ο,1 ]分別轉換成位址位元[A 2 2 ’,A 2 1 ’ ]= [〇,〇 ]、[ Ο,1 ]與[1 , Ο ]。易言之,位址轉換器4 1 d加1至位 址位元[A 2 2,A 2 1 ]上。 2 4位元位址信號a 〇 〇 - A 2 3可管理約1 6百萬數元組之位 址空間’並且半導體記憶裝置41 a/4 lb提供12百萬數元組 之資料儲存。在此情況中,除了位址位元 [A23,A22,A21,A20-A00] = [1,1,1,*,*,…*,*]與[ΐ,ι,〇,. t *,…,*,* ]外之所有記憶單元皆係可定址。因此,位址 位元A 2 3僅用於半導體記憶裝置4 1 a與4 1 b間之選擇性啟 動’並且如圖1 〇中所示位址位元A 2! 3、A 2 2、A 2 1剩餘的六 種組合係選擇性地指定至記憶單位陣列4丨e / 4丨g。微處理 器(未顯示出)並不傳送具有[A23,A22,A21] = [I,〆。 與[1,1,Ο ]之位址信號至位址匯流排4 2。然而,假統 設計者需要使用此等位址,則位址解碼器4 1 c會設計成在 此等位址出現下迫使資料輸出埠0UT4進入高阻抗狀態。當 記憶系統4 1更進一步地包含1 6百萬位元半導體裝置時,系 統設計者會需要使用此等位址。 從前述說明將瞭解,位址解碼器4 1 c與位址轉換器4 1 d、 相互合作以選擇性地指定位址次空間至半導體記憶裝置 41a/41b ’該二個半導體記憶裝置皆歸類為2的η次方以下Page 26 425575 V. Description of the invention (23) 4 lq and inverter 4 lp supply address bits A 22 ′ and A 2Γ to the address port AP4 of the semiconductor memory device 41b. The function of the address converter 41 d is expressed as A22 '= A22 · A 21 + A 22 · A21, A21' = A 21. Therefore, the address converter 4 1 d is the address bit [A 2 2, A 2 1] = [1,1], [〇, 〇] and [Ο, 1] are converted into address bits [A 2 2 ', A 2 1'] = [〇, 〇], [0,1 ] And [1, Ο]. In other words, the address converter 4 1 d adds 1 to the address bits [A 2 2, A 2 1]. The 24-bit address signal a 〇 〇-A 2 3 can manage an address space of about 16 million bytes' and the semiconductor memory device 41 a / 4 lb provides 12 million bytes of data storage. In this case, except for address bits [A23, A22, A21, A20-A00] = [1,1,1, *, *, ... *, *] and [ΐ, ι, 〇 ,. t *, …, *, *] Are all addressable. Therefore, the address bit A 2 3 is only used for selective activation of the semiconductor memory devices 4 1 a and 4 1 b 'and the address bits A 2! 3, A 2 2, A are shown in FIG. 10 2 The remaining six combinations are selectively assigned to the memory unit array 4 丨 e / 4 丨 g. The microprocessor (not shown) does not transmit data with [A23, A22, A21] = [I, 〆. With the address signal of [1, 1, 0] to the address bus 4 2. However, the pseudo-designer needs to use these addresses, and the address decoder 4 1 c will be designed to force the data output port OUT4 into a high impedance state when these addresses appear. When the memory system 41 further includes 16 million-bit semiconductor devices, the system designer will need to use these addresses. As will be understood from the foregoing description, the address decoder 4 1 c and the address converter 4 1 d cooperate with each other to selectively assign the address subspace to the semiconductor memory devices 41a / 41b 'the two semiconductor memory devices are classified Below 2 to the power of η

第27頁 -42557 5,,___ 五、發明說明(24) 的資料儲存容量。微處理器使位址信號A 0 0 - A 2 3從[0,0,0, *木***]依序地增力口至[1,〇,1,******]而不具有不可定址的 記憶單元。即使1 6位元程控指令碼儲存於半導體記憶裝置 4 1 a / 4 1 b中,但因為位址並未不連續,所以程式順序不會 受分支指令例如跳越至位址所擾亂。 半導體記憶裝置例如1 6百萬數元組半導體記憶裝置之 部分缺陷產品係可用於記憶系統,並且記憶系統之生產成 本會減低。 在此例子中,位址解碼器4 1 c與位址轉換器4 1 d整體構 成位址空間管理器。 半導體記憶裝置 第一實施例 參照圖1 1 ,實施本發明之半導體唯讀記憶裝置係製成 於半導體晶片5 0上。半導體唯讀記憶裝置包含記憶庫 5 1 / 5 2,並且記憶庫5 1 / 5 2提供關於位址空間之資料儲存 器°記憶庫5 1具有用以儲存3 C數元組之可定址記憶單元, 並且另一記憶庫5 2具有用以儲存1 C數元組、2 C數元組、3 C 數元組與4C數元組之記憶單元。C係表示成2的η次方之整 數。因此之故,記憶庫5 1之資料儲存容量少於4 C數元組。 位址空間中之高階位址係指定至記憶庫5 1 ,而低階位址則 指定至另一記憶庫5 2。 半導體唯讀記憶裝置進一步地包含位址解碼器5 3、可 程式位址轉換器54與資料缓衝器55。外部位址位元A 00-Αχ 或A 0 0 — A χ + 1係依據記憶庫5 2之資料儲存容量而供應至位址Page 27 -42557 5 ,, ___ V. The data storage capacity of invention description (24). The microprocessor makes the address signals A 0 0-A 2 3 sequentially increase from [0, 0, 0, * wood ***] to [1, 0, 1, ****] and Does not have non-addressable memory cells. Even if the 16-bit program-controlled instruction code is stored in the semiconductor memory device 4 1 a / 4 1 b, because the address is not discontinuous, the program sequence will not be disturbed by branch instructions such as jumping to the address. Some defective products of semiconductor memory devices such as 16 million-tuple semiconductor memory devices can be used in memory systems, and the production cost of memory systems will be reduced. In this example, the address decoder 4 1 c and the address converter 4 1 d integrally constitute an address space manager. Semiconductor Memory Device First Embodiment Referring to FIG. 11, a semiconductor read-only memory device embodying the present invention is fabricated on a semiconductor wafer 50. The semiconductor read-only memory device includes a memory bank 5 1/5 2 and the memory bank 5 1/5 2 provides a data storage about the address space. The memory bank 5 1 has an addressable memory unit for storing 3 C tuples. And, another memory bank 52 has a memory unit for storing a 1 C tuple, a 2 C tuple, a 3 C tuple, and a 4C tuple. C is expressed as an integer of 2 to the power of n. For this reason, the data storage capacity of memory 51 is less than 4 C bytes. Higher-order addresses in the address space are assigned to bank 5 1, and lower-order addresses are assigned to another bank 5 2. The semiconductor read-only memory device further includes an address decoder 5 3, a programmable address converter 54 and a data buffer 55. The external address bits A 00-Αχ or A 0 0 — A χ + 1 are supplied to the address according to the data storage capacity of the memory 5 2

第28頁 425575 五、發明說明(25) 解瑪器5 3 ’並且外部位址位元a X - 1與a X係供應至可程式位 址轉換器5 4 ◊位址解碼器5 3係回應於外部位址位元 AOO-Ax/AOO-Ax+1以產生選擇信號SEL1與内部位址位元。 選擇信號SEL1選擇性地使記憶庫5 1 / 5 2回應於内部位址位 元’並且資料數元組從記憶單元中讀出,該記憶單元係由 位址等同於由内部位址位元表示之位址戶片指定。資料數元 組轉移至資料缓衝器5 5,隨後傳送至資料埠5 6。 可程式位址轉換器5 4係由晶片賦能信號CE所賦能,並 且使外部位址位元A X / A X - 1轉換成内部位址位元a x與 ax-1。可程式位址轉換器54包含NOR閘54a/54b、反相器 54c/54d 、互斥-NOR 閘 54e 、反相器54f/54g/54h 、選用電 r 晶體54j/54k/54m/54n/54p/54q 與反相器 54r/54s/54t/54u。NOR閘54a/54b係由低態有效位準之晶 片賦能信號CE所賦能,並且產生反相位址位元cΑχ與 CAx-1。反相器54c/54d分別地重新產生位址位元Αχ與Page 28 425575 V. Description of the invention (25) Decoder 5 3 'and the external address bits a X-1 and a X are supplied to the programmable address converter 5 4 ◊ Address decoder 5 3 series response The external address bits AOO-Ax / AOO-Ax + 1 are used to generate the selection signal SEL1 and the internal address bits. The selection signal SEL1 selectively causes the memory bank 5 1/5 2 to respond to the internal address bit 'and the data byte is read from the memory unit, which is represented by the address equivalent to the internal address bit The address of the apartment is specified. The data byte is transferred to the data buffer 5 5 and then transmitted to the data port 5 6. The programmable address converter 54 is enabled by the chip enable signal CE and converts the external address bits A X / A X-1 into the internal address bits a x and ax-1. Programmable address converter 54 includes NOR gates 54a / 54b, inverters 54c / 54d, mutex-NOR gates 54e, inverters 54f / 54g / 54h, and electric crystals 54j / 54k / 54m / 54n / 54p / 54q and inverter 54r / 54s / 54t / 54u. The NOR gates 54a / 54b are energized by the low-level active-level wafer enable signal CE, and generate out-of-phase address bits cAχ and CAx-1. The inverters 54c / 54d respectively regenerate the address bits Aχ and

Ax -1 ° 依據是否通道摻雜而使選用電晶體54 j-54q作為半導 體熔線元件,並且提供導電通道。選用電晶體5 4卜5 4 q # 選擇性地通道摻雜如下。 μ 假使記憶庫5 2提供關於1 C數元組之資料儲存,糾切 ^ 則選用 電晶體54m/5 4q皆導通以提供導電通道,並且其他選 π 晶體54j/54k/54n/54p 皆關閉。 $ 則選用 用電晶 假使記憶庫52提供關於2C數元組之資料健存 電晶體54k/54p皆導通以提供導電通道,而其他選Ax -1 ° depends on whether the channel is doped, so that the transistor 54 j-54q is selected as the semiconductor fuse element, and a conductive channel is provided. The transistor 5 4b 5 4 q # is selected as follows. μ If the memory bank 5 2 provides data storage for 1 C tuples, rectification is used. The transistors 54m / 5 4q are selected to provide conduction, and the other selected π crystals 54j / 54k / 54n / 54p are turned off. $ Is used. If the transistor 52 is used to provide information about 2C tuples, the transistors 54k / 54p are all turned on to provide conductive channels.

第29頁 425575Page 425575

五、發明說明(26) 體54j/54m/54n/54q 皆關閉。 假使記憶庫5 2提供關於3 C數元組之資料儲存 電晶體54n/54q皆導通以提供導電通道,而其他’則選用 體54j/54k/54m/54p皆關閉。 、選用電晶 則選用 用電晶 假使記憶庫5 2提供關於4 c數元組之資料儲存 電晶體54j/54p皆導通以提供導電通道’而其他,弯 體54k/54m/54n/54Q 皆關閉 。 、 當選用電晶體54m/5 4q提供導電通道時,函數會表示5. Description of the invention (26) The bodies 54j / 54m / 54n / 54q are all closed. If the memory bank 5 2 provides data storage for 3 C tuples, the transistors 54n / 54q are all turned on to provide a conductive channel, and the other ’s are selected, and the body 54j / 54k / 54m / 54p is turned off. 1. If the transistor is used, the transistor is used. If the memory bank 5 2 provides data storage about the 4 c tuples, the transistors 54j / 54p are all turned on to provide a conductive channel ’, while the others, the curved body 54k / 54m / 54n / 54Q are all closed. When the transistor 54m / 5 4q is used to provide a conductive channel, the function will indicate

Ax = A x · A x ~ 1 + A x · A x -1,ax'l = Ax-1 當選用電晶體54k/54p提供導電通道時,函數會表示 ax = Ax, ax-1 = Αχ-1 當選用電晶體54n/54q提供導電通道時,函數會表示 ax 二 Αχ · A X-1 + A x ·Αχ-1, ax-1 = Ax-1 當選用電晶體54j/54p提供導電通道時,函數會表示 ax = Ax, ax-1 = Αχ -1 可程式位址轉換器5 4係依據導通的選用電晶體 5 4 j - 5 4 q而指定位址次空間至記憶庫5 1 / 5 2。 當選用電晶體54m/54ti導通時,如圖1 2A中所示位址空 間會分隔成類似記憶系統之第一實施例,下文t將無進一 步之說明。當選用電晶體54n/54Q導通時,如M2C中所示Ax = A x · A x ~ 1 + A x · A x -1, ax'l = Ax-1 When a transistor 54k / 54p is used to provide a conductive channel, the function will indicate ax = Ax, ax-1 = Αχ- 1 When the transistor 54n / 54q is used to provide a conductive channel, the function will represent ax II Ax · A X-1 + A x · Αχ-1, ax-1 = Ax-1 When a transistor 54j / 54p is used to provide a conductive channel , The function will indicate ax = Ax, ax-1 = Αχ -1 Programmable address converter 5 4 is based on the selection of the transistor 5 4 j-5 4 q to specify the address subspace to the memory 5 1/5 2. When the transistor 54m / 54ti is selected to be turned on, the address space is divided into the first embodiment similar to the memory system as shown in FIG. 12A, which will not be described further below. When the transistor 54n / 54Q is turned on, as shown in M2C

第30頁 ;425575 五、發明說明(27) 一~~ 位址空間會分隔成類似記憶系統之第二實施例,而說明也 省略。 假設選用電晶體5 4 k / 5 4 P係導通。當外部位址位元 Ax + 1、Ax、Ax、l …、A00 使位址從[0, 〇, 〇, ** …**]依序地 增加至[0 ’ ^ , 1,* H ]時,位址解碼器5 3會選擇記憶庫 5 2 ’並且資料數元組會從其中讀出(參看圖丨2 b )。在外 部位址位元Αχ+1、Αχ、Αχ-1···、Α0ϋ 超過[0,0,1,**...*木] 後:位址解碼器5 3會選擇另一記憶庫5丨,並且在外部位址 位元 Αχ+1、Αχ 'Ax-ι …、Α00 介於[〇,1,〇,**...**]與 [1,〇 ^ ^]間之期間中’資料數元組會從其中讀出。 、,,,假使選用電晶體5 4 j / 5 4 p導通,則位址解碼器5 3會基 於&quot;〇&quot;之外部位址位元Ax + 1而選擇記憶庫52,並且資料$ 兀組會從介於[0, 〇, **,…,**]至[丨,丨,&quot;,,**]間之外部位 f==Ax/Ax'i所特定之記憶單元中讀出(參看圖l4)。 η面^當外部位址位元^+丨以^^-1使位址從 满、U合且…依序地增加至[1,1,〇, **.·.**]時,位址解 瑪M3 了 $於&quot;丨,,之外部位址位元Αχ + ι而選擇另—記憶庫 並且貧料數兀組會從介於[0 , 〇,以…]至 [1,0,* *…木氺]間之内部位糾相- ’ ’ 6„β」間之内#位址位TLax、ax] .....a00所指 疋之&amp;己憶平元中讀出。 Η IΪ式位址轉換器5 4與位址解碼器5 3整體構成位址空、 ,並且N〇R閘54&amp;/5# '反相器54C/54d '互斥 X ^與反相器54卜54h級合形成邏輯電路。選用電晶 體54广54q分別提供複數個可斷裂之導電通道。Page 30; 425575 V. Description of the invention (27) 1 ~~ The address space is divided into a second embodiment similar to the memory system, and the description is omitted. It is assumed that the transistor 5 4 k / 5 4 P is selected to be turned on. When the external address bits Ax + 1, Ax, Ax, l…, A00 increase the address sequentially from [0, 〇, 〇, **… **] to [0 '^, 1, * H] At this time, the address decoder 5 3 will select the memory bank 5 2 ′ and the data tuples will be read from it (see figure 丨 2 b). After the external address bits Αχ + 1, Αχ, Αχ-1 ..., Α0ϋ exceed [0,0,1, ** ... * wood]: the address decoder 5 3 will select another memory bank 5丨, and in the period between the external address bits Aχ + 1, Aχ 'Ax-ι ..., A00 between [0, 1, 0, ** ... **] and [1, 0 ^ ^]' The data tuple is read from it. ,,,, If the transistor 5 4 j / 5 4 p is selected to be turned on, the address decoder 5 3 will select the memory bank 52 based on the &quot; 〇 &quot; outside address bit Ax + 1, and the data $ The group will read from the memory unit specified by [0, 〇, **, ..., **] to [丨, 丨, &quot; ,, **] and f == Ax / Ax'i Out (see Figure 14). η plane ^ When the external address bit ^ + 丨 makes the address from full, U, and ... sequentially increase to [1,1, 〇, **. ·. **] with ^^-1 The address solution M3 is $ Yu &quot; 丨, and the other locations are Αχ + ι and the other is selected—the memory bank and the number of poor data sets will be from [0, 〇, to ...] to [1, 0 , * * ... Mu 氺] internal bit phase correction-'' 6 „β" 内 内 #Address bit TLax, ax] ..... a00 refers to &amp; . The I-type address converter 5 4 and the address decoder 5 3 form an address space as a whole, and the NOR gate 54 &amp; / 5 # 'inverter 54C / 54d' mutually exclusive X ^ and inverter 54 Bu 54h cascade to form a logic circuit. The electric crystals 54 to 54q are selected to provide a plurality of rupturable conductive channels, respectively.

425575 發明說明¢28) 因為製造業者在制^ ^ 元ΑχΗ - ΑχΜ、庠選^己憶單元後可程式化外部位址位 係,所以可程式位址部位址位元ax - ax~i間之關 分缺陷產品。位址係Ϊ ^器54可使用半導體記憶裝置之部 元之記憶庫5〗/52的'記拾續苗地指定至不具有不可定址記憶單 第二實施例 …元上。 參照圖1 3 ,實脉* 於半導體晶片60上。ί f明之另一半導體記憶裝置係製成 位址解碼器6 3、位址Μ體δ己憶裝置包含記憶庫6 1 / 6 2、 具有關㈣數元^=器64與㈣緩衝㈤5。記憶庫61 於1C數元组之資料錯儲存器,而另一記憶庫62具有關 ^ ^ ^ fM / fi 9 ',谇窃。在此例子中,唯讀記憶單元形 Β思 °立址解碼器6 3與資料緩衝器6 5係類似各 位址解碼器53與資料緩衝器55,而為簡化起見在下文中不 加入進一步之說明。第二實施例之特徵為位址轉換器6 4, 據此’ s兒明係集中於位址轉換器6 4上。 位址轉換器64係依據控制信號⑽丁丨“”〗而改變位址 指定。假設外部位址位元A 〇 〇 - Αχ係定義位址空間,並且位 址空間之四分之三係不同地指定至記憶庫6 j 。425575 Description of the invention ¢ 28) Because the manufacturer can program the external address bit system after making ^ ^ element ΑχΗ-ΑχΜ, select ^ self-memory unit, so the programmable address location address bits ax-ax ~ i Distinguish defective products. The address system 54 can use the memory bank 5 of the semiconductor memory device's 52/52 to designate the serial number to have no non-addressable memory list. Second embodiment ... Referring to FIG. 13, the solid pulse * is on the semiconductor wafer 60. The other semiconductor memory device of the Ming Dynasty is made of an address decoder 6 3. The address M body δ memory device includes a memory bank 6 1/6 2. It has a key element ^ = 64 and a buffer 5. The memory 61 is in the wrong memory of the 1C tuple, and the other memory 62 has the key ^ ^ ^ fM / fi 9 ', plagiarism. In this example, the read-only memory unit B is an address decoder 63 and the data buffer 65 are similar to the address decoder 53 and the data buffer 55. For the sake of simplicity, no further description is added below. . The second embodiment is characterized by the address converter 64, and accordingly, the address converter 64 is focused on the address converter 64. The address converter 64 changes the address designation in accordance with the control signal "". It is assumed that the external address bits A 0-Αχ define an address space, and three-quarters of the address space are assigned differently to the memory bank 6 j.

位址轉換器64包含反相器64a/64b/64c/64d與NAND閘 64e/64f/64g/64h/64j/64k/64m/64n/64p/64Q。反相器 64a/64b產生外部位址位元Ax/Ax-1之反相信號,而反相器. 64c/64d產生控制信號OPT1/OPT2之反相信號。外部位址位 元Αχ/Ax-1、其反相信號、控制信號OPT 1/OPT2與其反相信 號係選擇.性地供應至NAND閘64e-64h與64k-64p,並且NANDThe address converter 64 includes inverters 64a / 64b / 64c / 64d and NAND gates 64e / 64f / 64g / 64h / 64j / 64k / 64m / 64n / 64p / 64Q. The inverters 64a / 64b generate the inverted signals of the external address bits Ax / Ax-1, and the inverters 64c / 64d generate the inverted signals of the control signals OPT1 / OPT2. The external address bits Αχ / Ax-1, its inversion signal, control signal OPT 1 / OPT2, and its anti-trust signal are selected. They are supplied to the NAND gates 64e-64h and 64k-64p, and NAND

第32頁 ^ 425575 五、發明說明(29) 問64e-64h之輸出節點與NAND間64k_64p之輸出節點係分別 地連接至N A N D閑6 4 j之輸入節點與μ a N D閘6 4 q之輸入節點。 内部位址位元ax/ax —1係分別地產生於NAND閘64 j/64q之輸 出節點處。 位址轉換器64達成下列關於内部位址位元ax與以—1之 邏輯函數。 ax = A x · Α χΊ · 〇 ΡΤ1 · 0ΡΤ2 + Αχ-1 · 0ΡΤ1 · 0ΡΤ2 + Αχ ·Αχ-1 + Αχ · 〇ΡΤ2 ax-Ι - Αχ-1 . 〇 ΡΤ1 · 〇 ΡΤ2 + α χΊ · ό ΡΤ1 · ΟΡΤ2 + Α χ ·Αχ-1 ·ΟΡΤ1 + Αχ · Α χ-1 ·ΟΡΤ1 此處0 ΡΤ1與ό ΡΤ2分別為ΟΡΤ1之反相信號與〇ΡΤ2之反相信丫 號。邏輯函數如圖1 4 Α中所示改變内部位址位元a χ並且如 圖1 4B中所示改變另一内部位址位元a χ - 1 。 如圖1 5 A所示,倘若控制信號Ο P T 1 / 〇 p τ 2係[〇,〇 ],則 從[0,0,**·_**]至[1,G,**..**]之外部位址位元會使位址 解碼器63與位址轉換器64指定從[0, 0, **..**]經 [0,1,**..**]至[1,〇,* *. - * * ]之位址次空間至記憶庫6 1 ., 並且剩餘的外部位址位元[1,1,**_.**]使位址解碼器6 3與 位址轉換器6 4指定剩餘的位址次空間至另一記憶庫6 2。如 圖15B所示,倘若控制信號0PT1/0PT2係[〇, ;!],則從[〇, 1,、 本至[1,1,木木*氺]之外部位址位元會使位址解碼器 63與位址轉換器64指定從[0,0,**.·**]經[〇,1,**.**]至— [1, 〇,**..**]之位址次空間至記憶庫6 1 ,並且剩餘的外部 位址位元[0,0,**..**]使位址解碼器6 3與位址轉換器6 4指Page 32 ^ 425575 V. Description of the invention (29) Question 64e-64h output node and 64k_64p output node between NAND are connected to the input node of NAND idle 6 4 j and μ a ND gate 6 4 q input node . The internal address bits ax / ax — 1 are generated at the output nodes of the NAND gate 64 j / 64q, respectively. The address converter 64 performs the following logical functions on the internal address bits ax and -1. ax = A x · ΑχΊ · 〇ΡΤ1 · 0ΡΤ2 + Αχ-1 · 0ΡΤ1 · 0ΡΤ2 + Αχ · Αχ-1 + Αχ · 〇ΡΤ2 ax-Ι-Αχ-1. 〇ΡΤ1 · 〇ΡΤ2 + α χΊ · ό ΤΤ1 · ΟΡΤ2 + Α χ · Αχ-1 · ΟΡΤ1 + Αχ · Αχ-1 · ΟΡΤ1 where 0 PT1 and ό PT2 are the reverse signals of ATP1 and TP2, respectively. The logic function changes the internal address bit a χ as shown in FIG. 14A and changes the other internal address bit a χ-1 as shown in FIG. 14B. As shown in FIG. 15A, if the control signal 0 PT 1 / 〇p τ 2 is [〇, 〇], then from [0,0, ** · _ **] to [1, G, ** ... **] Address bits outside the address will cause the address decoder 63 and the address converter 64 to specify from [0, 0, ** .. **] via [0, 1, ** .. **] to [1, 0, * *.-* *] Address space to the memory 6 1., And the remaining external address bits [1, 1, ** _. **] make the address decoder 6 3 and the address converter 6 4 designate the remaining address subspace to another memory 6 2. As shown in FIG. 15B, if the control signal 0PT1 / 0PT2 is [〇,;! ], Then from [〇, 1 ,, to [1, 1, Mumu * 氺], the address bits will cause the address decoder 63 and the address converter 64 to specify from [0, 0, ** . · **] via [〇, 1, **. **] to — [1, 〇, ** .. **] to the memory space 6 1 and the remaining external address bits [0, 0, ** .. **] make the address decoder 6 3 and the address converter 6 4

第33頁 4 2 5 5 7 5&quot; ---------- ----------! 五、發明說明(30) 定剩餘的位址次空間至另一記憶庫6 2。如圖1 5 C所示,倘 若控制信號OPT1 /OPT2係[ii ]二則[〇, 〇, *木.· **]與從 [1,0,**..**]至[1 , 1 ,**·.**]之外部位址位元會使位址解 碼器63與位址轉換器64指定[〇,〇,**._**]與從 [0,1,**.·**]至[1,〇,**..**]之位址次空間至記憶庫61 , 並且剩餘的外部位址位元[〇,1,**.,**]使位址解碼器6 3與 位址轉換器6 4指定剩餘的位址次空間至另一記憶庫6 2。如 圖1 5 D所示,倘若控制信號ορή / ορτ 2係[1,〇 ],則從[〇,0, 木*_ _**]至[0, 1,*木,與[1,1,**..**]之外部位址位元 會使位址解碼器6 3與位址轉換器6 4指定從[〇,〇 ,**·.**]至 [0,1,**..**]與[1,0,**·.**]之位址次空間至記憶庫6 1,一 並且剩餘的外部位址位元[1,0,**..**]使位址解碼器6 3與 位址轉換器6 4指定剩餘的位址次空間至另一記憶庫6 2。 因此,製造業者藉著使用控制信號OPT 1/0PT2而自由 地指定位址空間至記憶庫6 1 / 6 2。即使關於記憶庫6 2之位 址次空間使關於記憶庫6 1之位址次空間分隔為二,位址仍 然連續,並且記憶庫6 1 / 6 2不具有任何不可定址的記憶單 元。 第三實施例 參照圖1 6,實施本發明之又一半導體記憶裝置係製成 +於半導體晶片70上。半導體記憶裝置包含記憶庫71/72、 位址解碼器7 3、位址轉換器7 4以及資料緩衝器7 5。記憶庫 7 1 / 7 2、位址解碼器7 3與資料缓衝器7 5之動作分別類似記 憶庫6 1 / 6 2、位址.解碼器6 3與資料緩衝器6 5。因此,下文Page 33 4 2 5 5 7 5 &quot; ---------- ----------! V. Description of the invention (30) Set the remaining address space to another memory bank 6 2. As shown in FIG. 15C, if the control signal OPT1 / OPT2 is [ii], then two [〇, 〇, * 木. · **] and from [1, 0, ** .. **] to [1, 1, ** ·. **] causes the address decoder 63 and the address converter 64 to specify [〇, 〇, ** ._ **] and [0,1, ** . · **] to [1, 0, ** .. **] to the memory space 61, and the remaining external address bits [0, 1, **., **] enable The address decoder 6 3 and the address converter 6 4 designate the remaining address subspace to another bank 6 2. As shown in Figure 1D, if the control signal ορή / ορτ 2 is [1, 0], then from [0, 0, wood * _ _ **] to [0, 1, * wood, and [1, 1 , ** .. **] outside the address bit will cause the address decoder 63 and address converter 6 4 to specify from [〇, 〇, ** ·. **] to [0,1, * * .. **] and [1, 0, ** ·. **] address space to memory 6 1, one and the remaining external address bits [1, 0, ** .. ** ] Make the address decoder 6 3 and the address converter 6 4 assign the remaining address subspaces to another bank 6 2. Therefore, the manufacturer freely assigns the address space to the memory bank 6 1/62 by using the control signal OPT 1 / 0PT2. Even if the address space on memory 62 separates the address space on memory 62 into two, the addresses are continuous, and memory 6 1/62 does not have any non-addressable memory cells. Third Embodiment Referring to FIG. 16, another semiconductor memory device embodying the present invention is fabricated on a semiconductor wafer 70. The semiconductor memory device includes a memory bank 71/72, an address decoder 7 3, an address converter 74, and a data buffer 75. Memory bank 7 1/7 2. The operations of the address decoder 7 3 and the data buffer 75 are similar to those of the memory bank 6 1/62 2. The address. Decoder 6 3 and the data buffer 65. So below

第34頁 -/¾ P557 5_ 五、發明說明(31) 中將不說明此等組件7 1 / 7 2 / 7 3 / 7 5以避免重複。 位址轉換器7 4類似位址轉換器6 4係依據控制信號 Ο P T 1 / Ο P T 2而改變從夕卜部位址位元A X / A X - 1至内部位址位元 ax/ax-Ι之位址轉換。圖14A/14B與圖15A-15D係有關第三 實施例。使用互斥-0 R閘,邏輯閘間之配線會較位址轉換 器6 4之配線簡化。位址解碼器7 3與位址轉換器7 4整體構成 位址空間管理器,並且位址空間管理器達成位址空間管理 器6 3 / 6 4之全部優點。 從前述說明將可瞭解,位址空間管理器指定連續位址 空間之一部份至具有2的η次方以下之資料儲存容量而無不 可定址記憶單元之半導體記憶裝置。 即使半導體記憶裝置之產品經檢測為部分缺陷,該產 品仍可用於具有2的η次方以下之資料儲存容量之半導體記 憶裝置,而增加產品良率。 當關於微處理器之程式順序儲存於記憶庫或記憶系統 時,位址空間管理器會指定位址次空間至不具有不可定址 記憶單元之該記憶庫或該記憶系統上,並且任何指令例如 跳越皆不會擾亂程式順序之執行° .位址管理器6 3 / 6 4允許系統管理者指定連續位址次空 間之一部份至不同類型的記憶裝置,並且系統管理者可藉 著使用位址空間管理器而建構獨一的記憶系統。、Page 34-/ ¾ P557 5_ 5. The description of the invention (31) will not explain these components 7 1/7 2/7 3/7 5 to avoid repetition. The address converter 7 4 is similar to the address converter 6 4 according to the control signal 0 PT 1 / 〇 PT 2 and changes from the address bit AX / AX-1 to the internal address bit ax / ax-1. Address translation. 14A / 14B are related to the third embodiment of Figs. 15A-15D. With the mutex-0 R gate, the wiring between the logic gates is simplified compared to the wiring of the address converter 64. The address decoder 7 3 and the address converter 7 4 integrally constitute an address space manager, and the address space manager achieves all the advantages of the address space manager 6 3/6 4. As will be understood from the foregoing description, the address space manager specifies a portion of the continuous address space to a semiconductor memory device having a data storage capacity of 2 to the power of n without all addressable memory cells. Even if the product of the semiconductor memory device is detected as a partial defect, the product can still be used in a semiconductor memory device with a data storage capacity of less than 2 η to increase the product yield. When a program about a microprocessor is sequentially stored in a memory bank or a memory system, the address space manager assigns an address subspace to the memory bank or the memory system without an unaddressable memory unit, and any instruction such as jump More and more will not disturb the execution of the program sequence. The address manager 6 3/6 4 allows the system manager to specify a part of the continuous address sub-space to different types of memory devices, and the system manager can use the bit Address space manager to build a unique memory system. ,

雖然本發明之特別實施例業已顯示並說明,但是那些 熟習此項技術之人士明瞭得在不偏離本發明之精神與範圍 下製作各種改變與修正DAlthough specific embodiments of the present invention have been shown and described, those skilled in the art will understand that various changes and modifications can be made without departing from the spirit and scope of the present invention.

第35頁 425575 五、發明說明(32) 舉例言之,位址轉換器與半導體記憶器得分別分離地 積集於半導體晶片上。 半導體記憶裝置3 1 a / 3 1 b、位址解碼器3 1 c與位址轉 換器3 1 d得積集於半導體晶片上。相似地,半導體記憶裝 置4 1 a / 4 1 b、位址解碼器4 1 c與位址轉換器4 1 d得積集於半 導體晶片上。 關於選擇信號SEL1之位址解碼器53之一部份及/或可 程式位址轉換器5 4得與相當於記憶庫5 2 / 5 1之半導體記憶 裝置分離而積集於半導體晶片上。 相似地,位址解碼器6 3之一部份與位址轉換器6 4得與 相當於記憶庫61/62之半導體記憶裝置分離而積集於半導 γ 體晶片上。記憶庫6 2得與記憶庫6 1和位址空間管理器 6 3 / 6 4分離而積集於半導體晶片上。在此例子中,記憶庫 6 2之記憶單元之類型得不同於記憶庫6 1之記憶單元。假使 記憶庫6 1係唯讀記憶器,系統設計者得指定剩餘的位址次, 空間至替代唯讀記憶器之隨機存取記憶器或可電除且可程 式唯讀記憶器。 在半導體記憶裝置之第二與第三實施例以及半導體記 憶裝置之第一實施例中,其中記憶庫5 2提供關於1 C數元組 之資料儲存,位址空間管理器指定位址空間之四分之三至 記憶庫5 1 / 6 1 / 7 1 ,據此,需要二條外部位址信號線。然 而,假使位址空間管理器使位址空間分隔成8而用以指定Μ (自然數)/ 8至一個位址庫,則需要三條外部位址線。假 使位址空間分隔成1 6,則需要四條位址線。因此,連接至Page 35 425575 V. Description of the invention (32) For example, the address converter and the semiconductor memory must be separately and separately integrated on the semiconductor wafer. The semiconductor memory devices 3 1 a / 3 1 b, the address decoder 3 1 c and the address converter 3 1 d are accumulated on the semiconductor wafer. Similarly, the semiconductor memory device 4 1 a / 4 1 b, the address decoder 4 1 c and the address converter 4 1 d are integrated on a semiconductor wafer. A part of the address decoder 53 of the selection signal SEL1 and / or the programmable address converter 54 can be separated from the semiconductor memory device corresponding to the memory bank 5 2/51 and accumulated on the semiconductor chip. Similarly, a part of the address decoder 63 and the address converter 64 can be separated from the semiconductor memory device corresponding to the memory bank 61/62 and accumulated on the semiconductor gamma chip. The memory bank 6 2 must be separated from the memory bank 61 and the address space manager 6 3/6 4 and accumulated on the semiconductor wafer. In this example, the type of the memory cells in the memory bank 62 is different from the memory cells in the memory bank 61. If the memory bank 61 is a read-only memory, the system designer may specify the remaining address times and space to a random access memory or a programmable read-only memory that replaces the read-only memory. In the second and third embodiments of the semiconductor memory device and the first embodiment of the semiconductor memory device, the memory bank 52 provides data storage about 1 C tuples, and the address space manager specifies four of the address space. Three-thirds to the memory bank 5 1/6 1/7 1. Based on this, two external address signal lines are required. However, if the address space manager divides the address space into 8 to specify M (natural number) / 8 to an address library, three external address lines are required. If the address space is divided into 16 then four address lines are needed. So connect to

第36頁 ^-4255 7 5 五、發明說明(33) 位址轉換器之外部位址線係依據2的η次方以下之資料儲存 容量相對於位址空間2η之比例而變化。 在前述之實施例中,最低位址係零,亦即[〇,〇,0,…, 〇]。無論如何,最低位址絕不僅限於零。在依據本發明之 記憶系統或依據本發明之半導體記憶裝置中,最低位址得 為2的η次方,亦即2,4,8…中之任一個。 部份位址空間得指定至不同種類之資料儲存器例如硬 碟或介面。Page 36 ^ -4255 7 5 V. Description of the invention (33) The address lines outside the address converter are changed according to the ratio of the data storage capacity below 2 to the η position relative to the 2η address space. In the foregoing embodiment, the lowest address is zero, that is, [0, 0, 0, ..., 0]. In any case, the lowest address is by no means limited to zero. In the memory system according to the present invention or the semiconductor memory device according to the present invention, the lowest address can be n-th power of 2, which is any of 2, 4, 8,.... Part of the address space may be assigned to different types of data storage such as hard drives or interfaces.

第37頁Page 37

Claims (1)

^ 4 2 5 5 7 5 六、申請專利範圍 1 . 一種位址空間管理器,其指定由(η 1 + η 2 )個位址位 元所定義之位址空間的位址次空間至資料儲存器(3 1 b ; 41b ;51 ;61 ;71),該資料儲存器用以儲存m 1區段資料 資訊,此處該η 1 、該η 2與該m 1係第一自然數、第二自然數 與第三自然數,並且該ml係等於或者大於且小於2nl ; 其特徵為包含: 位址轉換器(3 1 d ; 4 1 d ; 5 4 ; 6 4 ; 7 4 ),其轉換該η 1 個位址位元(A 2 1 / A 2 2 ; A X / A X - 1 ),並且供應經轉換之位 址位元(A 2 1 ’/ A 2 2 ’; a X / a X - 1 )至該資料儲存器以使該位 址次空間定位於該位址空間中2nl_x個位置之一上,此處該X 係滿足方程式2nl-ml = 2X ;以及 位址解碼器(3 1 c / 3 2 ; 4 1 c / 4 2 ; 5 3 ; 6 3 ; 7 3 ),其從 包含該η 2個位址位元之位址信號產生解碼信號 (ΑΟ0-Α20 ; aO-ax-2 ),並且供應該解碼信號至該資料儲 存器以便從儲存於該位址次空間中之該m 1區段資料資訊中 選取次區段資料資訊。 2. 如申請專利範圍第1項之位址空間管理器,其中該 位址次空間包含一連串位址,該一連_位址係連續於該位 址空間中剩餘位址次空間之位址。 3. 如申請專利範圍第1項之位址空間管理器,其中該 位址轉換器包含: 邏輯電路(5 4 a - 5 4 h ),被供應以該η 1個位:位元 (Ax/Ax- 1 )·並且從該η 1個位址位元產生該經轉換之位址 位元(a X / a X - 1 )之候選者,以及^ 4 2 5 5 7 5 6. Scope of patent application 1. An address space manager that specifies the address subspace of the address space defined by (η 1 + η 2) address bits to data storage (3 1 b; 41b; 51; 61; 71), the data storage is used to store the data information of the m 1 section, where the η 1, the η 2 and the m 1 are the first natural number and the second natural Number and the third natural number, and the ml is equal to or greater than and less than 2nl; it is characterized by including: an address converter (3 1 d; 4 1 d; 5 4; 6 4; 7 4), which converts the η 1 address bit (A 2 1 / A 2 2; AX / AX-1), and supply the converted address bit (A 2 1 '/ A 2 2'; a X / a X-1) To the data storage so that the address subspace is positioned on one of the 2nl_x positions in the address space, where the X system satisfies the equation 2nl-ml = 2X; and the address decoder (3 1 c / 3 2; 4 1 c / 4 2; 5 3; 6 3; 7 3), which generates a decoded signal (ΑΟ0-Α20; aO-ax-2) from an address signal containing the n 2 address bits, and Supply the decoded signal to the data storage to Select sub section data feed information from the data stored in the address m times of a space section. 2. For example, the address space manager of the first scope of the patent application, wherein the address subspace includes a series of addresses, and the consecutive _addresses are addresses that are consecutive to the remaining address subspaces in the address space. 3. For example, the address space manager of the first patent application range, wherein the address converter includes: a logic circuit (5 4 a-5 4 h), which is supplied with the η 1 bit: bit (Ax / Ax-1), and a candidate for the converted address bit (aX / aX-1) is generated from the n1 address bits, and 第38頁 42557 5 々、申請專利範圍 複數個可斷裂導電通道(5 4 j - 5 4 q ),其分別關聯於 該候選者,並且選擇性地導通以供應該經轉換的位址位元 至該資料儲存器。 4. 如申請專利範圍第3項之位址空間管理器,其中該 複數個可斷裂導電通道係分別形成相互並聯之複數個電晶 體(54j-54q)。 5. 如申請專利範圍第4項之位址空間管理器,其中該 複數個電晶體(5 4 j - 5 4 q )係選擇性地形成正常開類型以 便從該等候選者中選擇該經轉換之位址位元。 6. 如申請專利範圍第5項之位址空間管理器,其中正 常開類型電晶體(5 4 j - 5 4p )係經由在其製造過程中所實 現之通道摻雜而決定。 7. 如申請專利範圍第3項之位址空間管理器,其中該 複數個可斷裂導電通道(5 4 j - 5 4q )係選擇性地導通以使 在該位址次空間中之一連串位址連續於在該位址空間的剩 餘位址次空間中之位址。 8. 如申請專利範圍第1項之位址空間管理器,其中該 位址轉換器(64 ;74)包含邏輯電路(64a-64q),該邏 輯電路回應於表示該2 nl_x個位置中之一之該nl個位址位元 (Ax/Ax-1 )與控制信號(0PT1/0PT2 ),以便在複數個範 圍中之一決定該經轉換之位址位元(a X / a X - 1 ),該複數 個範圍中之一係分-別對應於該2nl—x個位置中之一。 9. 如申請專利範圍第8項之位址空間管理器,其中該 控制信號(0PT1 /0PT2 )改變該2nl_x個位置中之一之表示。Page 38 42557 5 々. The scope of the patent application is a plurality of rupturable conductive channels (5 4 j-5 4 q), which are respectively associated with the candidate and are selectively turned on to supply the converted address bits to The data store. 4. The address space manager of item 3 of the patent application, wherein the plurality of rupturable conductive channels respectively form a plurality of electric crystals (54j-54q) connected in parallel with each other. 5. The address space manager of item 4 in the patent application scope, wherein the plurality of transistors (5 4 j-5 4 q) are selectively formed into a normal on type to select the converted from the candidates Address bit. 6. The address space manager of item 5 in the scope of patent application, in which normally-on transistors (5 4 j-5 4p) are determined by channel doping achieved during the manufacturing process. 7. The address space manager of item 3 of the patent application scope, wherein the plurality of rupturable conductive channels (5 4 j-5 4q) are selectively turned on to make a series of addresses in the address subspace. Consecutive addresses in the remaining address subspaces of that address space. 8. If the address space manager of item 1 of the patent application scope, wherein the address converter (64; 74) contains a logic circuit (64a-64q), the logic circuit responds to indicate one of the 2 nl_x positions The nl address bits (Ax / Ax-1) and the control signal (0PT1 / 0PT2) to determine the converted address bits (a X / a X-1) in one of a plurality of ranges One of the plurality of ranges corresponds to-one corresponding to each of the 2nl-x positions. 9. The address space manager of item 8 in the patent application scope, wherein the control signal (0PT1 / 0PT2) changes the representation of one of the 2nl_x positions. 第39頁 Zl 2 5 5 7 5 __ 六、申請專利範圍 10. 如申請專利範圍第1項之位址空間管理器,其中 半導體記憶裝置(3 1 b ; 4 1 b )係作為該資料儲存器。 11. 如申請專利範圍第1 0項之位址空間管理器,其中 該半導體記憶裝置(3 1 b ; 4 1 b )具有可定址的唯讀記憶單 元。 12. 如申請專利範圍第1項之位址空間管理器,其中 該位址解碼器(3 1 c ; 4 1 c ; 53 ; 6 3 ; 7 3 )係選擇該資料儲 存器與受該位址空間之另一位址次空間所指定之另一資料 儲存器其中之一。 13. 如申請專利範圍第1 2項之位址空間管理器,其中 該另一資料儲存器係用以儲存m2區段資料資訊,此處m2係 第四自然數,並且該ml與該m2之和係等於或者大於2nH且 小於2ηί。 1 4 . 如申請專利範圍第1 3項之位址空間管理器,其中 一連串位址與另一連串位址皆分別併入該位址次空間與該 另一位址次空間中,並且該一連串位址係連續於該另一連 串位址。 15. 如·申請專利範圍第1 3項之位址空間管理器,其中 該資料儲存器(5 1 ; 6 1 ; 7 1 )與該另一資料儲存器(5 2 ; 6 2 ; 7 2 )皆為一種半導體記憶裝置。 16. 如申請專利範圍第1 5項之位址空間管理器,其中 該種半導體記憶裝置係唯讀記憶裝置。 17. 如申請專利範圍第1 3項之位址空間管理器,其中 該資料儲存器(6 1 )與該另一資料儲存器(6 2 )分別為不Page 39 Zl 2 5 5 7 5 __ VI. Patent application scope 10. For example, the address space manager of the first patent application scope, wherein the semiconductor memory device (3 1 b; 4 1 b) is used as the data storage device. . 11. The address space manager of item 10 in the patent application scope, wherein the semiconductor memory device (31b; 41b) has an addressable read-only memory unit. 12. If the address space manager of item 1 of the patent application scope, wherein the address decoder (3 1 c; 4 1 c; 53; 6 3; 7 3) select the data storage and receive the address One of the other data storage designated by the space. 13. For example, the address space manager of item 12 in the patent application scope, wherein the other data storage is used to store the data information of the m2 section, where m2 is the fourth natural number, and the ml and the m2 The sum is equal to or greater than 2nH and less than 2ηί. 14. If the address space manager of item 13 of the scope of patent application is applied, one series of addresses and another series of addresses are respectively merged into the address subspace and the other address subspace, and the series of bits The address is consecutive to the other series of addresses. 15. For example, the address space manager of the 13th scope of the patent application, wherein the data storage (5 1; 6 1; 7 1) and the other data storage (5 2; 6 2; 7 2) All are a kind of semiconductor memory device. 16. For example, the address space manager of item 15 of the patent application scope, wherein the semiconductor memory device is a read-only memory device. 17. If the address space manager of item 13 of the patent application scope, wherein the data storage (6 1) and the other data storage (6 2) are not 第40頁 d255T5 六、申請專利範圍 同類型之半導體記憶裝置。 18. 如申請專利範圍第1 7項之位址空間管理器,其中 該不同類型之半導體記憶裝置中之一係唯讀記憶裝置。 19. 一種半導體記憶裝置,受供應以定義位址空間之 .(η 1 + η 2 )個位址位元,此處該η 1與該η 2係第一自然數與 第二自然數,包含: 一個以上之記憶庫(5 1 / 5 2 ; 6 1 / 6 2 ; 7 1 / 7 2 ),其中 之一 (5 1 ; 6 1 ; 7 1 )係儲存m 1區段資料資訊於該位址空間 之位址次空間中,此處該m 1係第三自然數並且等於或者大 於2⑷-丨)且小於2n】;以及 用以產生内部位址信號(a 0 - a X )之位址空間管理 為 , 其特徵為: 該位址空間管理器包含: 位址轉換器(5 4 ; 6 4 ; 7 4 ),其用以轉換該η 1個位址 位元,並且供應經轉換之位址位元(a X / a X - 1 )至該一個 以上記憶庫中之一,以使該位址次空間定位於該位址空間 中2nl_x個位置中之一,此處該X滿足方程式2nI-m 1二2X,以 及 位址解碼器(5 3 ; 6 3 ; 7 3 ),從包含至少該η 2個位址 位元之位址信號(ΑΟΟ-Αχ+1 ; ΑΟΟ-Αχ )產生選擇信號 (SEL1 )與解碼信號,以該選擇信號選擇地啟動該一個以 上之記憶庫,並且供應該解碼信號至該資料儲存器以便從 儲存於該位址次空間中之m 1區段資料資訊中選取次區段資Page 40 d255T5 VI. Scope of Patent Application Semiconductor memory devices of the same type. 18. The address space manager of item 17 in the patent application scope, wherein one of the different types of semiconductor memory devices is a read-only memory device. 19. A semiconductor memory device that is supplied with (η 1 + η 2) address bits defining an address space, where η 1 and η 2 are a first natural number and a second natural number, including : More than one memory bank (5 1/5 2; 6 1/6 2; 7 1/7 2), one of which (5 1; 6 1; 7 1) stores m 1 section data information in the bit In the address space of the address space, here m 1 is the third natural number and is equal to or greater than 2⑷- 丨) and less than 2n]; and the address used to generate the internal address signal (a 0-a X) The space management is characterized by: The address space manager includes: an address converter (5 4; 6 4; 7 4), which is used to convert the n 1 address bits and supply the converted bits Address bit (a X / a X-1) to one of the one or more memories, so that the address subspace is positioned at one of the 2nl_x positions in the address space, where X satisfies the equation 2nI -m 1 2 2X, and an address decoder (5 3; 6 3; 7 3), generates a selection from an address signal (ΑΟΟ-Αχ + 1; ΑΟΟ-Αχ) containing at least the η 2 address bits. Signal (SEL1) and decoded signal, selectively activate the one or more memory banks with the selection signal, and supply the decoded signal to the data storage so as to extract from the m 1 section data information stored in the address subspace Select subsection 第41頁 4 2 5 5 7 5 六,申請專利範圍 料資訊。 20. 如申請專利範圍第1 9項之半導體記憶裝置,其中 該位址次空間包含一連串位址,該一連串位址係連續於該 位址空間中剩餘位址次空間之位址,該剩餘位址次空間之 位址係指定至該一個以上記憶庫之另一個。 2 1.如申請專利範圍第2 0項之半導體記憶裝置,其中 該位址轉換器(5 4 )包含 邏輯電路(5 4 a - 5 4 h ),受供應以該η 1個位址位元並 且從該η 1個位址位元產生該經轉換之位址位元之候選者, 以及 複數個可斷裂導電通道(5 4 j - 5 4 q ),其分別關聯於 該候選者,並且選擇性地導通以供應該經轉換之位址位元 至該資料儲存器上。 22. 如申請專利範圍第2 1項之半導體記憶裝置,其中 該複數個可斷裂導電通道係分別形成相互並聯之複數個電 晶體(5 4卜5 4 q )。 23. 如申請專利範圍第2 2項之半導體記憶裝置,其中 該複數個電晶體(5 4 j - 5 4 q )係選擇性地形成正常開類型 以便從該候選者中選出該經轉換之位址位元,並且正常開 類型電晶體係經由在其製造過程中所實現之通道摻雜而決 定。 24. 如申請專利範圍第1 9項之半導體記憶裝置,其中 該位址轉換器(64 ; 74 )係回應於表示該2nI_x個位置中之 —之控制信號(OPT1 /OPT2 ),以便在複數個範圍中之一Page 41 4 2 5 5 7 5 VI. Patent application scope Material information. 20. For a semiconductor memory device according to item 19 of the patent application, wherein the address subspace includes a series of addresses, the series of addresses are consecutive addresses in the remaining address subspace in the address space, and the remaining bits The address of the address space is assigned to another one of the more than one memory bank. 2 1. The semiconductor memory device according to item 20 of the patent application scope, wherein the address converter (5 4) includes a logic circuit (5 4 a-5 4 h), and is supplied with the η 1 address bit And the candidate of the converted address bit and the plurality of rupturable conductive channels (5 4 j-5 4 q) are generated from the n 1 address bits, which are respectively associated with the candidate, and are selected It is conductively supplied to supply the converted address bits to the data storage. 22. The semiconductor memory device of claim 21, wherein the plurality of rupturable conductive channels respectively form a plurality of transistors (54 4 5 4 q) connected in parallel with each other. 23. The semiconductor memory device as claimed in claim 22, wherein the plurality of transistors (5 4 j-5 4 q) are selectively formed into a normal on type so as to select the converted bit from the candidate. Address bit, and the normal on-type transistor system is determined by channel doping achieved during its manufacturing process. 24. The semiconductor memory device of item 19 in the scope of patent application, wherein the address converter (64; 74) is in response to a control signal (OPT1 / OPT2) indicating one of the 2nI_x positions so that One of the range 第42頁 六、申請專利範圍 決定該經轉換之位址位元,該複數個範圍中之一係分別對 應於該2 η 1 - X個位置中之一。 25. 如申請專利範圍第1 9項之半導體記憶裝置,其中 該一個以上之記憶庫中之一具有可定址唯讀記憶單元。 2 6. —種記憶系統,包含: 複數個記憶單位(3 1 a / 3 1 b ; 4 1 a / 4 1 b ),其分別由位 址空間之位址次空間所指定而用以儲存資料資訊並且回 應於位址信號(A0〇-A22/A00-A20/A21’/A22’ )而用以至 少讀出該資料資訊,該複數個記憶單位中之一儲存m 1區段 資料資訊,此處m 1係大於2的(η - 1 )次方且小於2的η次方 並且η係自然數;以及 位址空間管理器,其回應於該位址信號之較高的位址 位元(A 2 1- A 2 2 ; A 2 1 - A 2 3 )而用以選擇性地賦能該複數個 記憶單位並且選擇性地轉換該較的位址位元以修正形成 該位址信號之一部份之位址位元(A 2 Γ / A 2 2 ’)而替代該 較高的位址位元; 其特徵為: 該位址空間管理器(3 1 c / 3 1 d ; 4 1 c / 4 1 d )使該較高的 位址位元轉換成該修正位址位元,而以在該複數個記憶單 位中不具有不可定址記憶單元之方式i該位址次空間連 續。 27.如申請專利範圍第2 6項之記憶系統,其中該複數 個記憶單位之另一個係由具有以表示為K之最高位址的該 位址次空間中之另一個所.指定,並且該最高位址係連續於Page 42 6. Scope of patent application Determine the converted address bits. One of the plurality of ranges corresponds to one of the 2 η 1-X positions, respectively. 25. The semiconductor memory device of claim 19, wherein one of the more than one memory bank has an addressable read-only memory unit. 2 6. — A memory system, including: a plurality of memory units (3 1 a / 3 1 b; 4 1 a / 4 1 b), which are respectively designated by the address subspace of the address space to store data Information and responds to the address signal (A0〇-A22 / A00-A20 / A21 '/ A22') to read at least the data information, one of the plurality of memory units stores m 1 section data information, this Where m 1 is a power of (η-1) greater than 2 and a power of η less than 2 and η is a natural number; and an address space manager that responds to a higher address bit of the address signal ( A 2 1- A 2 2; A 2 1-A 2 3) to selectively enable the plurality of memory units and selectively convert the lower address bits to modify and form one of the address signals Part of the address bit (A 2 Γ / A 2 2 ') instead of the higher address bit; its characteristics are: the address space manager (3 1 c / 3 1 d; 4 1 c / 4 1 d) The higher address bit is converted into the modified address bit, and the address subspace is connected in such a manner that there is no unaddressable memory unit in the plurality of memory units. Continued. 27. The memory system according to item 26 of the patent application scope, wherein the other one of the plurality of memory units is designated by the other one of the sub-spaces of the address having the highest address indicated as K, and the The highest address is consecutive 425575 六、申請專利範圍 以表示為(K + 1 )之該位址次空間中之該一個的最低位 址° 28. 如申請專利範圍第2 7項之記憶系統,其中該位址 空間管理器將該較高位址位元之值減去該K,用以決定該 修正位址位元。 29. 如申請專利範圍第2 7項之記憶系統,其中該記憶 單位之該另一個係儲存m2區段資料資訊,並且該m2係2的k 次方,此處k係自然數。 30. 如申請專利範圍第2 7項之記憶系統,其中該記憶 單位之該另一個係儲存m 2區段資料資訊,並且該m 2係大於 2的(j - 1 )次方且小於2的j次方並且]係自然數。425575 6. The scope of the patent application is expressed as (K + 1) the lowest address of the one in the address subspace. 28. For the memory system of scope 27 of the patent application, the address space manager The value of the higher address bit is subtracted from the K to determine the modified address bit. 29. For example, the memory system of item 27 in the scope of patent application, wherein the other unit of the memory unit stores m2 section data information, and the m2 is a k-th power of 2, where k is a natural number. 30. For example, the memory system of item 27 in the scope of patent application, wherein the other unit of the memory unit stores m 2 section data information, and the m 2 is greater than 2 (j-1) to the power of less than 2 jth power] is a natural number. 第44頁Page 44
TW088108696A 1998-05-28 1999-05-25 Address space manager for assigning part of address space to data storage without non-addresable cell, semiconductor memory device with built-in address space manager and memory system TW425575B (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP14715798A JPH11338765A (en) 1998-05-28 1998-05-28 Address translating circuit and semiconductor memory device

Publications (1)

Publication Number Publication Date
TW425575B true TW425575B (en) 2001-03-11

Family

ID=15423883

Family Applications (1)

Application Number Title Priority Date Filing Date
TW088108696A TW425575B (en) 1998-05-28 1999-05-25 Address space manager for assigning part of address space to data storage without non-addresable cell, semiconductor memory device with built-in address space manager and memory system

Country Status (3)

Country Link
JP (1) JPH11338765A (en)
KR (1) KR19990088620A (en)
TW (1) TW425575B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI450119B (en) * 2010-01-13 2014-08-21 Htc Corp Method for addressing management object in management tree and associated device management system

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100735024B1 (en) * 2005-12-29 2007-07-03 삼성전자주식회사 An address converter of a semiconductor device and semiconductor memory device
US10579290B2 (en) 2016-03-23 2020-03-03 Winbond Electronics Corp. Option code providing circuit and providing method thereof

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI450119B (en) * 2010-01-13 2014-08-21 Htc Corp Method for addressing management object in management tree and associated device management system

Also Published As

Publication number Publication date
JPH11338765A (en) 1999-12-10
KR19990088620A (en) 1999-12-27

Similar Documents

Publication Publication Date Title
TW497031B (en) Synchronous content addressable memory with single cycle operation
US6521994B1 (en) Multi-chip module having content addressable memory
US6324114B1 (en) Semiconductor memory device using a plurality of semiconductor memory chips mounted in one system and a semiconductor memory system using a plurality of semiconductor memory devices
KR19990088601A (en) Programmable pin designation for semiconductor devices
JPH02139797A (en) Variable field content reading memory
JPS5998262A (en) Memory access method and bidirectional data byte array device
US6718432B1 (en) Method and apparatus for transparent cascading of multiple content addressable memory devices
US20130322183A1 (en) Semiconductor device and semiconductor memory device
JPH01821A (en) Priority encoder
JP2000353388A (en) Improvement of contents referable memory
TWI509618B (en) Packaged spi-nand flash memory device and flash memory device and configuring method therof
TW425575B (en) Address space manager for assigning part of address space to data storage without non-addresable cell, semiconductor memory device with built-in address space manager and memory system
EP0358773B1 (en) Microcomputer
JPS5843832B2 (en) memory device
CN112037829B (en) Enable signal generating circuit and semiconductor device using the same
US7102925B2 (en) Flash memory device
KR100391524B1 (en) Semiconductor memory device with replacement programming circuit
JPH03254499A (en) Semiconductor storage device
US4975882A (en) User programmable redundant memory
TW440830B (en) Redundant form address decoder for memory system
US11100267B1 (en) Multi dimensional memory compression using bytewide write enable
US7085181B2 (en) Semiconductor device having storage circuit which stores data in nonvolatile manner by using fuse element
KR100821583B1 (en) Circuit and method for controlling redundancy in semiconductor memory apparatus
US5233561A (en) Composite semiconductor storage device and operating method therefor
TWI235380B (en) Output multiplexing implementation for a simultaneous operation flash memory device

Legal Events

Date Code Title Description
GD4A Issue of patent certificate for granted invention patent
MM4A Annulment or lapse of patent due to non-payment of fees