Semiconductor storage Number:7,024,524 from the United States Patent and Trademark Office (PTO) owispatent

Sponsored Links

 
Web LinkGrinder.com

Title: Semiconductor storage

Abstract: It is an object to obtain a semiconductor storage having a 1—chip structure which can be simultaneously accessed to memory cells present in different memory cell arrays. A 1-port memory cell array (11) provided with a word line (WL1) for a first port in common and a 2-port memory cell array (12) are provided together over one chip, thereby constituting a semiconductor storage. By selectively bringing any of a plurality of the word lines (WL1) for the first port into an active state by a row decoder (16), it is possible to simultaneously access respective memory cells of the 1-port memory cell array (11) and the 2-port memory cell array (12). By selectively bringing any of a plurality of word lines (WL2) for a second port into an active state by a row decoder (18), it is possible to singly access the 2-port memory cell array (12).

Patent Number: 7,024,524 Issued on 04/04/2006 to Nii


Inventors: Nii; Koji (Tokyo, JP)
Assignee: Renesas Technology Corp. (Tokyo, JP)
Appl. No.: 454500
Filed: June 5, 2003
Foreign Application Priority Data

Dec 10, 2002[JP]2002-357630

Current U.S. Class: 711/154; 711/100; 711/117; 711/123; 711/136
Current Intern'l Class: G06F 12/00    (20060101)
Field of Search: 711/100,111,131,149,150,154,156,117,118,123,125,136 365/189.01,233,239


References Cited [Referenced By]

U.S. Patent Documents
4964083Oct., 1990Nogle et al.
5067111Nov., 1991Asano et al.
6144055Nov., 2000Takenaka.
6169684Jan., 2001Takahashi et al.
6195278Feb., 2001Calin et al.
6229758May., 2001Agata.
Foreign Patent Documents
6-349275Dec., 1994JP.
10-178110Jun., 1998JP.

Primary Examiner: Thai; Tuan V.
Attorney, Agent or Firm: Oblon, Spivak, McClelland, Maier & Neustadt, P.C.

Claims



What is claimed is:

1. A semiconductor storage comprising first and second memory cell arrays,

said first memory cell array including:

a plurality of first memory cells arranged in at least one column over a plurality of rows; and

a plurality of first word lines to be connected to said first memory cells on a row unit,

said second memory cell array including:

a plurality of second memory cells arranged in a matrix, said second memory cells including a different number of elements than said first memory cells;

a plurality of second word lines to be connected to said second memory cells on a row unit;

a plurality of third word lines connected to said second memory cells on a row unit and not connected to any of said plurality of first memory cells;

a plurality of first bit lines provided corresponding to said second memory cells on a column unit and being accessible to said second memory cell connected to said second word line in a selection state out of said plurality of second word lines; and

a plurality of second bit lines provided corresponding to said second memory cells on a column unit and being accessible to said second memory cell connected to said third word line in a selection state out of said plurality of third word lines,

said semiconductor storage further comprising:

a first row decoder configured to select, so that a period wherein any of said first word lines is brought into a selection state based on a first address signal and a period wherein any of said second word lines is brought into a selection state based on said first address signal are overlapped.

2. The semiconductor storage according to claim 1, wherein

said first memory cells include a plurality of memory cells arranged in a matrix; and

said first and second word lines include a plurality of common word line to be shared on a row unit.

3. The semiconductor storage according to claim 1, wherein

said first memory cells include a plurality of memory cells arranged in a matrix; and

said first row decoder is provided between said first and second memory cell arrays.

4. The semiconductor storage according to claim 1, wherein

said first and second memory cells include circuit structures having different port structures from each other.

5. The semiconductor storage according to claim 4, wherein

said first memory cell includes a memory cell having a 1-port structure which can be accessed by only a first port,

said second memory cell includes a memory cell having a 2-port structure which can be accessed by said first port and a second port, and

an access to be carried out by said first row decoder in a selection state of said first and second word lines includes an access to be preformed by said first port.

6. The semiconductor storage according to claim 5, further comprising:

a second row decoder for bringing any of said plurality of said third word lines into a selection state based on a second address signal,

an access to be carried out by said second row decoder in a selection state of said third word line including an access to be performed by said second port.

7. The semiconductor storage according to claim 5, wherein

said second memory cell includes a memory cell having a 2-port structure in which reading and writing can be carried out by said second port.

8. The semiconductor storage according to claim 5, wherein

said second memory cell includes a content addressable memory cell in which a comparison result between an expected value given from an outside and self-stored contents can be output from said second port.

9. The semiconductor storage according to claim 5, wherein

said second memory cell includes a memory cell having a 2-port structure in which only reading using said second port can be carried out.

10. The semiconductor storage according to claim 5, wherein

said first memory cell and said second memory cell have cell heights to be formation lengths in a predetermined direction on a layout structure which are set to be equal to each other.

11. The semiconductor storage according to claim 4, further comprising:

a second row decoder for bringing any of said plurality of said third word lines into a selection state based on a second address signal, wherein

said first memory cell includes a memory cell having a 2-port structure which can be accessed by a first port and a second port,

said second memory cell including a memory cell having a 2-port structure which can be accessed by said first port and a third port which is different from said second port,

an access to be carried out by setting said first and second word lines into a selection state by said first row decoder including an access to be performed by said first port, and

an access to be carried out by setting said third word line into a selection state by said second decoder including an access to be performed by said third port.

12. The semiconductor storage according to claim 1, wherein

said plurality of first memory cells include a plurality of first dummy cells and said plurality of second memory cells include a plurality of memory cells for a real operation,

said plurality of first and second word lines include a plurality of common word lines to be shared on a row unit, and

the number of accessible ports of said first dummy cells is set to be smaller than that of accessible ports of said memory cells for a real operation.

13. The semiconductor storage according to claim 12, further comprising:

a second row decoder for bringing any of said plurality of third word lines into a selection state based on a second address signal; and

a third memory cell array having a plurality of second dummy cells arranged in at least a column over a plurality of rows, said first and third memory cell arrays being formed integrally with said second memory cell array interposed therebetween and the number of accessible ports of said second dummy cell being set to be smaller than that of accessible ports of said memory cells for a real operation,

said memory cells for a real operation including a memory cell having a 2-port structure with first and second ports which are accessible,

said first and second dummy cells including a memory cell having a 1-port structure with first and second dummy ports which are accessible,

said first ports of said plurality of memory cells for a real operation and said first dummy ports of said plurality of first dummy cells being connected to said plurality of common word lines on a row unit in common,

said second ports of said plurality of memory cells for a real operation and said second dummy ports of said plurality of second dummy cells being connected to said plurality of third word lines on a row unit in common,

said first row decoder including a row decoder provided in the vicinity of said first memory cell array, and

said second row decoder including a row decoder provided in the vicinity of said third memory cell array.

14. The semiconductor storage according to claim 13, wherein

said first and second dummy cells include a memory cell for reading timing regulation with respect to said memory cell for a real operation.

15. A semiconductor storage comprising first and second memory cell arrays,

said first memory cell array including:

a plurality of first memory cells arranged in at least one column over a plurality of rows; and

a plurality of first word lines to be connected to said first memory cells on a row unit,

said second memory cell array including:

a plurality of second memory cells arranged in a matrix;

a plurality of second word lines to be connected to said second memory cells on a row unit;

a plurality of third word lines connected to said second memory cells on a row unit and not connected to any of said plurality of first memory cells;

a plurality of first bit lines provided corresponding to said second memory cells on a column unit and being accessible to said second memory cell connected to said second word line in a selection state out of said plurality of second word lines; and

a plurality of second bit lines provided corresponding to said second memory cells on a column unit and being accessible to said second memory cell connected to said third word line in a selection state out of said plurality of third word lines,

said semiconductor storage further comprising:

a first row decoder for simultaneously bringing any of said first word lines and any of said second word lines into a selection state based on a first address signal;

a second row decoder for bringing any of said plurality of third word lines into a selection state based on a second address signal; and

a third memory cell array having a plurality of second dummy cells arranged in at least a column over a plurality of rows, said first and third memory cell arrays being formed integrally with said second memory cell array interposed therebetween and the number of accessible ports of said second dummy cell being set to be smaller than that of accessible ports of said memory cells for a real operation,

said memory cells for a real operation including a memory cell having a 2-port structure with first and second ports which are accessible,

said first and second dummy cells including a memory cell having a 1-port structure with first and second dummy ports which are accessible,

said first ports of said plurality of memory cells for a real operation and said first dummy ports of said plurality of first dummy cells being connected to said plurality of common word lines on a row unit in common,

said second ports of said plurality of memory cells for a real operation and said second dummy ports of said plurality of second dummy cells being connected to said plurality of third word lines on a row unit in common,

said first row decoder including a row decoder provided in the vicinity of said first memory cell array, and

said second row decoder including a row decoder provided in the vicinity of said third memory cell array.

16. The semiconductor storage according to claim 15, wherein

said first and second dummy cells include a memory cell for reading timing regulation with respect to said memory cell for a real operation.

17. A semiconductor storage comprising first and second memory cell arrays,

said first memory cell array including:

a plurality of first memory cells arranged in at least one column over a plurality of rows; and

a plurality of first word lines to be connected to said first memory cells on a row unit,

a plurality of second word lines to be connected to said first memory cells on a row unit,

a plurality of first bit lines provided corresponding to said first memory cells on a column unit and being accessible to said first memory cell connected to said first word line in a selection state out of said plurality of first word lines; and

a plurality of second bit lines provided corresponding to said first memory cells on a column unit and being accessible to said first memory cell connected to said second word line in a selection state out of said plurality of second word lines,

said second memory cell array including:

a plurality of second memory cells arranged in a matrix, said second word lines not being connected to any of said second memory cells;

a plurality of third word lines to be connected to said second memory cells on a row unit;

a plurality of fourth word lines connected to said second memory cells on a row unit and not connected to any of said first memory cells;

a plurality of third bit lines provided corresponding to said second memory cells on a column unit and being accessible to said second memory cell connected to said third word line in a selection state out of said plurality of third word lines; and

a plurality of fourth bit lines provided corresponding to said second memory cells on a column unit and being accessible to said second memory cell connected to said fourth word line in a selection state out of said plurality of fourth word lines,

said semiconductor storage further comprising:

a first row decoder configured to select so that a period wherein any of said first word lines is brought into a selection state based on a first address signal and a period wherein any of said third word lines is brought into a selection state based on said first address signal are overlapped,

a second row decoder configured to bring any of said of second word lines into a selection state based on a second address signal; and

a third row decoder configured to bring any of said of fourth word lines into a selection state based on a third address signal, wherein

said first memory cell includes a memory cell having a 2-port structure which can be accessed by a first port and a second port,

said second memory cell including a memory cell having a 2-port structure which can be accessed by said first port and a third port which is different from said second port,

an access to be carried out by setting said first and third word lines into a selection state by said first row decoder including an access to be preformed by said first port,

an access to be carried out by setting said second word line into a selection state by said second row decoder including an access to be performed by said second port, and

an access to be carried out by setting said fourth word line into a selection state by said third row decoder including an access to be performed by said third port.

18. The semiconductor storage according to claim 17, wherein

said second and third row decoders are arranged with said first and second memory cell arrays interposed therebetween.
Description



BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates to a semiconductor storage and more particularly to a combination of a plurality of memory cell structures of an MOS static RAM.

2. Description of the Background Art

For a semiconductor storage comprising plural kinds of memory cells having different port structures and the like, conventionally, a semiconductor storage has been disclosed in Patent Document 1 (Japanese Patent Application Laid-Open No. 6-349275 (1994) gazette), for example. The semiconductor storage comprises a 3-port cell section which is 3-port simultaneous accessible and a 1-port cell section having one access port, and can be implemented by connecting them to at least a pair of bit lines in common.

On the other hand, it has been demanded that data to be accessed on a bit length unit are divided on a several-bit unit and individual data are accessed on the several-bit unit thus obtained by the division in a computer field or the like. Moreover, it has also been demanded that a part of bits are constituted to be accessible from a plurality of ports.

In the semiconductor storage of the Patent Document 1 described above, however, the 3-port cell section and the 1-port cell section share a bit line. For this reason, there is a problem in that both of the port cell sections cannot be accessed at the same time and the demand cannot be met.

SUMMARY OF THE INVENTION

It is an object of the present invention to obtain a semiconductor storage having a one-chip structure in which memory cells present in different memory cell arrays can be accessed at the same time.

According to the present invention, a semiconductor storage includes first and second memory cell arrays and a first row decoder, and the first memory cell array includes a plurality of first memory cells and a plurality of first word lines. The plurality of first memory cells are arranged in at least one column over a plurality of rows and the plurality of first word lines are connected to the first memory cells on a row unit. On the other hand, the second memory cell array includes a plurality of second memory cells, a plurality of second and third word lines, and a plurality of first and second bit lines. The plurality of second memory cells are arranged in a matrix, the second word lines are connected to the second memory cells on a row unit, the plurality of third word lines are connected to the second memory cells on a row unit and are not connected to any of the plurality of first memory cells, and the plurality of first bit lines are provided corresponding to the second memory cells on a column unit and are accessible to the second memory cell connected to the second word line in a selection state out of the plurality of second word lines, and the plurality of second bit lines are provided corresponding to the second memory cells on a column unit and are accessible to the second memory cell connected to the third word line in a selection state out of the plurality of third word lines. The first row decoder simultaneously brings any of the first word lines and any of the second word lines into the selection state based on a first address signal.

The first and second memory cell arrays can be controlled by the first row decoder. Consequently, it is possible to obtain a semiconductor storage having a 1-chip structure which can be accessed to the first and second memory cells at the same time.

These and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram showing a structure of a semiconductor storage according to a first embodiment of the present invention,

FIG. 2 is a circuit diagram showing a memory cell array structure according to the first embodiment,

FIG. 3 is an explanatory view showing a layout structure provided under a first aluminum wiring layer in a 1-port memory cell as seen on a plane,

FIG. 4 is an explanatory view showing a layout structure provided above a second aluminum wiring layer in the 1-port memory cell,

FIG. 5 is an explanatory view showing a layout structure provided under a first aluminum wiring layer in a 2-port memory cell as seen on a plane,

FIG. 6 is an explanatory view showing a layout structure provided above the second aluminum wiring layer in the 2-port memory cell,

FIG. 7 is an explanatory view showing a layout structure provided under a first aluminum wiring layer in a 1-port and 2-port memory cell as seen on a plane,

FIG. 8 is a block diagram showing a structure of a semiconductor storage according to a second embodiment of the present invention,

FIG. 9 is a block diagram showing a structure of a semiconductor storage according to a third embodiment of the present invention,

FIG. 10 is an explanatory diagram showing an internal structure of a row decoder in FIG. 9,

FIG. 11 is a block diagram showing a structure of a semiconductor storage according to a fourth embodiment of the present invention,

FIG. 12 is a circuit diagram showing a memory cell array structure according to the fourth embodiment,

FIG. 13 is a circuit diagram showing the details of a CAM memory cell,

FIG. 14 is an explanatory view showing a layout structure of all layers in the CAM memory cell as seen on a plane,

FIG. 15 is an explanatory view showing a layout structure provided under a first aluminum wiring layer in FIG. 14 as seen on a plane,

FIG. 16 is an explanatory view showing a layout structure provided above a second aluminum wiring layer in FIG. 14 as seen on a plane,

FIG. 17 is a block diagram showing a structure of a semiconductor storage according to a fifth embodiment of the present invention,

FIG. 18 is a circuit diagram showing a memory cell array structure according to the fifth embodiment,

FIG. 19 is a block diagram showing a structure of a semiconductor storage according to a sixth embodiment of the present invention,

FIG. 20 is a circuit diagram showing a structure of a 2-port memory cell according to the sixth embodiment,

FIG. 21 is an explanatory view showing a layout structure of all layers of the 2-port memory cell according to the sixth embodiment as seen on a plane,

FIG. 22 is an explanatory view showing a layout structure provided under a first aluminum wiring layer in FIG. 21 as seen on a plane,

FIG. 23 is an explanatory view showing a layout structure provided above the first aluminum wiring layer in FIG. 21 as seen on a plane,

FIG. 24 is a block diagram showing a structure of a semiconductor storage according to a seventh embodiment of the present invention,

FIG. 25 is a circuit diagram showing a structure of a dummy cell formed in a memory cell area for timing regulation, and

FIG. 26 is a circuit diagram showing a memory cell array structure according to the seventh embodiment.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

<First Embodiment>

(Whole Structure)

FIG. 1 is a block diagram showing a structure of a semiconductor storage according to a first embodiment of the present invention, As shown in FIG. 1, a 1-port memory cell array 11 (a first memory cell array) and a 2-port memory cell array 12 (a second memory cell array) are provided together over one chip to constitute a memory macro. More specifically, the 1-port memory cell array 11 is provided with a word line WL1 for a first port (a first word line (a common word line)) and the 2-port memory cell array 12 is provided with a word line WL1 for a first port (a second word line (a common word line)) and a word line WL2 for a second port (a third word line).

Upon receipt of an address input bus signal AD1, a control circuit 31 supplies a row address to a row decoder 16 (a first row decoder) and a column address to column selectors 21 and 22 under timing control of a read control input signal RE1 and a write control input signal WE1.

Upon receipt of an address input bus signal AD2, a control circuit 32 supplies a row address to a row decoder 17 (a second row decoder) and a column address to a column selector 23 under timing control of a read control input signal RE2 and a write control input signal WE2.

The row decoder 16 selectively sets any of a plurality of the word lines WL1 for the first port into an active state (a selection state) based on the row address and the row decoder 17 selectively sets any of a plurality of the word lines WL2 for the second port into the active state based on the row address.

The column selector 21 selects a portion equivalent to 4 bits from a plurality of bit line pairs BL1 and bar BL1 for the first port based on the column address sent from the control circuit 31 and inputs/outputs a data input/output bus signal DIO11<7:4> to/from the 1-port memory cell array 11 through the bit line pair BL1 and bar BL1 for the first port thus selected.

The column selector 22 selects a portion equivalent to 4 bits from a plurality of bit line pairs BL21 and bar BL21 for the first port (first bit lines) based on the column address sent from the control circuit 31 and inputs/outputs a data input/output bus signal DIO12<3:0> to/from the 2-port memory cell array 12 through the bit line pairs BL21 and bar BL21 for the first port thus selected.

The column selector 23 selects a portion equivalent to 4 bits from a plurality of bit line pairs BL22 and bar BL22 for the second port (second bit lines) based on the column address sent from the control circuit 32 and inputs/outputs a data input/output bus signal DIO2<3:0> to/from the 2-port memory cell array 12 through the bit line pairs BL22 and bar BL22 for the second port thus selected.

These column selectors 22 to 24 have a sense amplifier (SA) function and a write driving (WD) function.

(Memory Cell Structure)

FIG. 2 is a circuit diagram showing a memory cell array structure of the 1-port memory cell array 11 and the 2-port memory cell array 12. For convenience of explanation, FIG. 2 shows only a 2×2 memory cell but does not imply an actual size of the memory cell array (a matrix-like memory cell arrangement).

The 1-port memory cell array 11 is constituted by memory cells ms00, ms01, ms10 and ms11.

Each of the memory cells ms00 to ms11 is constituted by inverters 41 and 42 which are cross connected to each other, an NMOS transistor Q11 having one of electrodes connected to an input of the inverter 41 (an output of the inverter 42), and an NMOS transistor Q12 having one of electrodes connected to an input of the inverter 42 (an output of the inverter 41).

Word lines WL1<0> and WL1<1> for the first port are driven by drivers 52 and 51 (which are usually provided in the row decoder 16 (not shown)), respectively.

In the memory cells ms00 and ms01 on the same row, gate electrodes of the NMOS transistors Q11 and Q12 are electrically connected to the word line WL1<0> for the first port in common. In the memory cells ms10 and ms11 on the same row, gate electrodes of the NMOS transistors Q11 and Q12 are electrically connected to the word line WL1<1> for the first port in common.

Both of the memory cells ms00 and ms10 on the same column are provided between a bit line pair BL1<0> and bar BL1<0> for the first port. In these memory cells ms00 and ms10, the other electrode of the NMOS transistor Q11 is electrically connected to the bit line BL1<0> for the first port and the other electrode of the NMOS transistor Q12 is electrically connected to the inverted bit line bar BL1<0> for the first port.

Both of the memory cells ms01 and ms11 on the same column are provided between a bit line pair BL1<1> and bar BL1<1> for the first port. In the memory cells ms01 and ms11, the other electrode of the NMOS transistor Q11 is electrically connected to the bit line BL1<1> for the first port and the other electrode of the NMOS transistor Q12 is electrically connected to the inverted bit line bar BL1<1> for the first port.

On the other hand, the 2-port memory cell array 12 is constituted by memory cells md00, md01, md10 and md11.

Each of the memory cells md00 to md11 is constituted by the inverters 41 and 42 which are cross connected to each other, NMOS transistors Q11 and Q21 having electrodes connected to the input of the inverter 41, and NMOS transistors Q12 and Q22 having electrodes connected to the input of the inverter 42.

Word lines WL2<0> and WL2<1> for the second port are driven by drivers 54 and 53 (which are usually provided in the row decoder 17 (not shown)), respectively.

In the memory cells md00 and md01 on the same row, gate electrodes of the NMOS transistors Q11 and Q12 are electrically connected to the word line WL1<0> for the first port in common, and gate electrodes of the NMOS transistors Q21 and Q22 are electrically connected to the word line WL2<0> for the second port in common.

In the memory cells md10 and md11 on the same row, gate electrodes of the NMOS transistors Q11 and Q12 are electrically connected to the word line WL1<1> for the first port in common, and gate electrodes of the NMOS transistors Q21 and Q22 are electrically connected to the word line WL2<1> for the second port in common.

Both of the memory cells md00 and md10 on the same column are provided between a bit line pair BL21<0> and bar BL21<0> for the first port. In these memory cells md00 and md10, the other electrode of the NMOS transistor Q11 is electrically connected to the bit line BL21<0> for the first port and the other electrode of the NMOS transistor Q12 is electrically connected to the inverted bit line bar BL21<0> for the first port.

Furthermore, both of the memory cells md00 and md10 are provided between a bit line pair BL22<0> and bar BL22<0> for the second port. In these memory cells md00 and md10, the other electrode of the NMOS transistor Q21 is electrically connected to the bit line BL22<0> for the second port and the other electrode of the NMOS transistor Q22 is electrically connected to the inverted bit line bar BL22<0> for the second port.

Both of the memory cells md01 and md11 on the same column are provided between a bit line pair BL21<1> and bar BL21<1> for the first port. In these memory cells md01 and md11, the other electrode of the NMOS transistor Q11 is electrically connected to the bit line BL21<1> for the first port and the other electrode of the NMOS transistor Q12 is electrically connected to the inverted bit line bar BL21<1> for the first port.

Furthermore, both of the memory cells md01 and md11 are provided between a bit line pair BL22<1> and bar BL22<1> for the second port. In these memory cells md10 and md11, the other electrode of the NMOS transistor Q21 is electrically connected to the bit line BL22<1> for the second port and the other electrode of the NMOS transistor Q22 is electrically connected to the inverted bit line bar BL22<1> for the second port.

In FIGS. 1 and 2, and FIG. 3 and subsequent drawings, a symbol < > represents a bus signal (line) and a numeral in < > corresponds to a bus signal name. Moreover, <n-1:0> represents a signal having an n-bit width of bus signals 0 to (n-1).

(Operation)

With reference to FIGS. 1 and 2, description will be given to a reading and writing operation using the first port and a reading and writing operation using the second port. For convenience of explanation, a memory cell array structure of the 1-port memory cell array 11 and the 2-port memory cell array 12 is set to be n (rows)×m (columns).

First of all, in the case in which the reading operation of the first port is to be executed, the read control input signal RE1 is enabled. When the read control input signal RE1 is brought into the enable state, the row decoder 16 for receiving a row address corresponding to the address input bus signal AD1 through the control circuit 31 selectively sets, into an active state, a word line WL1<i> (i=0 to n-1) for the first port in word lines WL1<n-1:0> for the first port having n bits.

Consequently, the NMOS transistors Q11 and Q12 of each of selected memory cells msi0 to msim of the 1-port memory cell array 11 connected to the word line WL1<i> for the first port which is set into the active state and selected memory cells mdi0 to mdim of the 2-port memory cell array 12 are turned ON, and contents stored in each of the selected memory cells are propagated to corresponding bit line pairs BL1 and bar BL1, and BL21 and bar BL21 for the first port.

Each of the column selectors 21 and 22 for receiving the column address corresponding to the address input bus signal AD1 through the control circuit 31 selects four sets of bit line pairs from m bit line pairs BL1 and bar BL1, and BL21 and bar BL21 for the first port, and senses and amplifies them by means of an internal sense amplifier. Consequently, the data input/output bus signal DIO11<7:4> and the data input/output bus signal DIO12<3:0> are output as 8-bit read data.

In the case in which the writing operation of the first port is to be executed, next, the write control input signal WE1 is enabled. When the write control input signal WE1 is brought into the enable state, the row decoder 16 for receiving a row address corresponding to the address input bus signal AD1 through the control circuit 31 selectively sets, into an active state, the word line WL1<i> (i=0 to n-1) for the first port in the word lines WL1<n-1:0> for the first port having n bits.

Consequently, the NMOS transistors Q11 and Q12 of each of the selected memory cells msi0 to msim of the 1-port memory cell array 11 and the selected memory cells mdi0 to mdim of the 2-port memory cell array 12 which are connected to the word line WL1<i> for the first port set into the active state are turned ON, and the inverters 41 and 42 of each selected memory cell are electrically connected to the bit line pairs BL1 and bar BL1, and BL21 and bar BL21 for the first port.

At the same time, write data sent from the outside are given as the data input/output bus signal DIO11<7:4> and the data input/output bus signal DIO12<3:0> to/from the column selectors 21 and 22.

Each of the column selectors 21 and 22 for receiving the column address corresponding to the address input bus signal AD1 through the control circuit 31 selects four sets of bit line pairs from m bit line pairs BL1 and bar BL1, and BL21 and bar BL21 for the first port, and gives the write data to the selected bit line pairs, thereby driving the selected bit line pairs to "L" or "H". As a result, it is possible to carry out writing based on the write data for the selected memory cell.

Then, in the case in which the reading operation of the second port is to be executed, the read control input signal RE2 is enabled. When the read control input signal RE2 is brought into the enable state, the row decoder 17 for receiving a row address corresponding to the address input bus signal AD2 through the control circuit 32 selectively sets, into an active state, a word line WL2<i> (i=0 to n-1) for the second port in word lines WL2<n-1:0> for the second port having n bits.

Consequently, the NMOS transistors Q21 and Q22 of each of selected memory cells mdi0 to mdim of the 2-port memory cell array 12 connected to the word line WL2<i> for the second port which is set into the active state are turned ON, and contents stored in each of the selected memory cells are propagated to corresponding bit line pairs BL22 and bar BL22 for the second port.

Subsequently, the column selector 23 for receiving a column address corresponding to the address input bus signal AD2 through the control circuit 32 selects four sets of bit line pairs from m bit line pairs BL22 and bar BL22 for the second port respectively, and senses and amplifies them by means of an internal sense amplifier. Consequently, the data input/output bus signal DIO2<3:0> are output as 4-bit read data.

In the case in which the writing operation of the second port is to be executed, next, the write control input signal WE2 is enabled. When the write control input signal WE2 is brought into the enable state, the row decoder 17 for receiving a row address corresponding to the address input bus signal AD2 through the control circuit 32 selectively sets, into an active state, the word line WL2<i> for the second port in the word lines WL2<n-1:0> for the second port having n bits.

Consequently, the NMOS transistors Q21 and Q22 of each of the selected memory cells mdi0 to mdim of the 2-port memory cell array 12 connected to the word line WL2<i> for the second port which is set into the active state are turned ON, and the inverters 41 and 42 of each selected memory cell are electrically connected to the bit line pair BL22 and bar BL22 for the second port.

At the same time, write data sent from the outside are given as the data input/output bus signal DIO2<3:0> to/from the column selector 23.

The column selector 23 for receiving a column address corresponding to the address input bus signal AD2 through the control circuit 32 selects four sets of bit line pairs from m bit line pairs BL22 and bar BL22 for the second port, and gives the write data to the selected bit line pairs, thereby driving the selected bit line pairs to "L" or "H". As a result, it is possible to carry out writing based on the write data for a selected memory cell.

As described above, it is possible to read and write 8-bit data which are divided and held in high and low orders on a 4-bit unit in the 1-port memory cell array 11 and the 2-port memory cell array 12 from the first port.

Then, it is possible to read and write the lower 4-bit data which are held in the 2-port memory cell array 12 from the second port. In this case, the upper 4-bit data are not accessed at all.

As a result, the semiconductor storage according to the first embodiment can divide, on a 4-bit unit, data to be accessed on an 8-bit length unit and can access data on a lower 4-bit unit thus obtained by the division. The lower 4-bit is accessible from the first and second ports so that it is possible to carry out reading and writing to meet a demand for a data dividing access.

In the first embodiment, thus, the 1-port memory cell array 11 and the 2-port memory cell array 12 including memory cells having different circuit structures from each other can be controlled simultaneously by the row decoder 16. Therefore, it is possible to obtain a semiconductor storage having a 1-chip structure in which a single port memory cell and a 2-port memory cell having different structures can be accessed simultaneously.

Even if the 1-port memory cell array 11 is constituted with a replacement by a 2-port memory cell, moreover, an equivalent operation to that of the first embodiment can be carried out. As compared with this case, an area can be decreased because the upper 4 bits are constituted by the 1-port memory cell, and a layout area can be decreased because it is possible to eliminate a control circuit, a column selector and the like for the upper 4 bits which do not need to be accessed in the second port. Since a circuit scale is thus decreased, it is possible to reduce power consumption during an operation and a standby.

(Layout Structure)

FIGS. 3 and 4 are explanatory views showing a layout structure of a single port memory cell MS of the 1-port memory cell array 11. FIG. 3 is an explanatory view mainly showing a layout structure provided under a first aluminum wiring layer as seen on a plane. FIG. 4 is an explanatory view showing a layout structure provided above a second aluminum wiring layer as seen on a plane.

It is assumed that the inverter 41 shown in FIG. 2 is a CMOS inverter constituted by a PMOS transistor P1 and an NMOS transistor N1 and the inverter 42 is a CMOS inverter constituted by a PMOS transistor P2 and an NMOS transistor N2.

As shown in FIG. 3, the PMOS transistors P1 and P2 are formed in an N well region NW, the NMOS transistor N1 and the NMOS transistor Q11 are formed in a P well region PW1, and the NMOS transistor N2 and the NMOS transistor Q12 are formed in a P well region PW0. The P well region PW0 and the P well region PW1 are formed opposite to each other with the N well region NW interposed therebetween.

In the N well region NW, the PMOS transistor P1 is constituted by a polysilicon wiring PL1 provided across a P+ diffusion region FLP1 and the PMOS transistor P2 is constituted by a polysilicon wiring PL2 provided across a P+ diffusion region FLP2.

In the P well region PW1, the NMOS transistor N1 is constituted by the polysilicon wiring PL1 provided across an N+ diffusion region FLN1 and the NMOS transistor Q11 is constituted by a polysilicon wiring PL11 provided across an N+ diffusion region FLQ11. The polysilicon wiring PL1 is formed from the N well region NW to the P well region PW1 and is thus shared as gates of the NMOS transistor N1 and the PMOS transistor P1.

In the P well region PW0, the NMOS transistor N2 is constituted by the polysilicon wiring PL2 provided across an N+ diffusion region FLN2 and the NMOS transistor Q12 is constituted by a polysilicon wiring PL12 provided across an N+ diffusion region FLQ12. The polysilicon wiring PL2 is formed from the N well region NW to the P well region PW0 and is thus shared as gates of the NMOS transistor N2 and the PMOS transistor P2.

Moreover, the polysilicon wiring PL1 and the polysilicon wiring PL12 are formed on the same straight line, the polysilicon wiring PL2 and the polysilicon wiring PL11 are formed on the same straight line, the diffusion regions FLP1, FLN1 and FLQ12 are formed on the same straight line in almost the same form, and the diffusion regions FLP2, FLN2 and FLQ11 are formed on the same straight line in almost the same form.

As a result, the PMOS transistor P1, the NMOS transistor N1 and the NMOS transistor Q12 can be formed along the same straight line and the PMOS transistor P2, the NMOS transistor N2 and the NMOS transistor Q11 can be formed along the same straight line. Consequently, a cell height HC1 of the single port memory cell MS can be set corresponding to two transistors. In this specification, the cell height implies a formation length in a direction of formation of a bit line (a longitudinal direction in the drawing) on the layout structure.

The P+ diffusion regions FLP1 and FLP2 are obtained by implanting and diffusing a P-type impurity and the N+ diffusion regions FLN1, FLN2, FLQ11 and FLQ12 are obtained by implanting and diffusing an N-type impurity. In the description of FIG. 3, each diffusion region has an upper region referred to as one of regions and a lower region referred to as the other region with respect to the polysilicon wiring in FIG. 3.

In the P well region PW1, a ground wiring LG1 (a first layer aluminum wiring) provided on one of the regions of the N+ diffusion region FLN1 is electrically connected to the diffusion region FLN1 through a diffusion contact hole CH. The polysilicon wiring PL11 is electrically connected to a word line 1WL1 (the first layer aluminum wiring) through a gate contact hole GC, and a bit line 1BL1 (the first layer aluminum wiring) provided on the other region of the diffusion region FLQ11 is electrically connected to the other region of the diffusion region FLQ11 through the diffusion contact hole CH.

The diffusion contract hole CH implies a contact hole of the diffusion region and the first layer (aluminum) wiring, and the gate contact hole GC implies a contact hole of the polysilicon wiring and the first layer wiring.

In the N well region NW, a power wiring LV1 (a first layer aluminum wiring) provided on one of the regions of the diffusion region FLP1 is electrically connected to one of the regions of the diffusion region FLP1 through a diffusion contact hole CH, and the power wiring LV1 provided on the other region of the diffusion region FLP2 is electrically connected to the other region of the diffusion region FLP2 through the diffusion contact hole CH.

In the P well region PW0, the inverted bit line bar BL1 (a first layer aluminum wiring) provided on one of the regions of the diffusion region FLQ12 is electrically connected to one of the regions of the diffusion region FLQ1 through a diffusion contact hole CH, and the word line 1WL1 provided on the polysilicon wiring PL12 is electrically connected to the polysilicon wiring PL12 through a gate contact hole GC. A ground wiring LG1 provided on the other region of the diffusion region FLN2 is electrically connected to the other region of the diffusion region FLN2 through a contact hole CH.

The aluminum wiring AL11 to be the first layer aluminum wiring extended from the other region of the diffusion region FLN1 to the other region of the P+ diffusion region FLP1 is electrically connected to one of the regions of the diffusion region FLN1 through the diffusion contact hole CH. The aluminum wiring AL11 is electrically connected to the other region of the P+ diffusion region FLP1 and the polysilicon wiring PL2 through a shared contact SC formed from the other region of the P+ diffusion region FLP1 to the polysilicon wiring PL2.

The aluminum wiring AL12 to be the first layer aluminum wiring extended from one of the regions of the diffusion region FLN2 to one of the regions of the P+ diffusion region FLP2 is electrically connected to one of the regions of the diffusion region FLN2 through the diffusion contact hole CH. The aluminum wiring AL12 is electrically connected to one of the regions of the P+ diffusion region FLP2 and the polysilicon wiring PL1 through the shared contact SC formed from one of the regions of the P+ diffusion region FLP2 to the polysilicon, wiring PL1.

As shown in FIG. 4, a ground wiring LG2 (a second layer aluminum wiring) is electrically connected to the lower ground wiring LG1 (not shown) through a via hole VH1. A power wiring LV2 (the second layer aluminum wiring) is electrically connected to the lower power wiring LV1 (not shown) through the via hole VH1.

A word line 2WL1 (the second layer aluminum wiring) is electrically connected to the word line 1WL1 (not shown) through the via hole VH1, and a word line 3WL1 (a third layer aluminum wiring) is electrically connected to the word line 2WL1 through a via hole VH2. These word lines 1WL1 to 3WL1 constitute the word line WL1 for the first port in FIGS. 1 and 2.

The word line 3WL1 is formed across the P well regions PW0 and PW1 and the N well region NW. The via hole VH1 implies a via hole for connecting the first layer wiring to the second layer (aluminum) wiring, and the via hole VH2 implies a via hole for connecting the second layer wiring to the third layer (aluminum) wiring.

The bit line 2BL1 (the second layer aluminum wiring) is electrically connected to the lower bit line 1BL1 (not shown) through the via hole VH1 and the inverted bit line bar 2BL1 (the second layer aluminum wiring) is electrically connected to the lower inverted bit line bar 1BL1 (not shown) through the via hole VH1.

The bit line pair BL1 and bar BL1 for the first port in FIGS. 1 and 2 are constituted by the bit line 2BL1, the bit line 1BL1 and the inverted bit lines bar 2BL1 and bar 1BL1.

The bit lines 2BL1 and bar 2BL1, the ground wiring LG2 and the power wiring LV2 are formed in parallel with each other in a longitudinal direction of FIG. 4 over the P well regions PW1 and PW0 and the N well region NW, respectively.

FIGS. 5 and 6 are explanatory views showing a layout structure of a 2-port memory cell MD of the 2-port memory cell array 12. FIG. 5 is an explanatory view mainly showing a layout structure provided under the first aluminum wiring layer as seen on a plane. FIG. 6 is an explanatory view showing a layout structure provided above the second aluminum wiring layer as seen on a plane.

It is assumed that the inverter 41 shown in FIG. 2 is a CMOS inverter constituted by a PMOS transistor P1 and an NMOS transistor N1 and the inverter 42 is a CMOS inverter constituted by a PMOS transistor P2 and an NMOS transistor N2.

As shown in FIG. 5, the PMOS transistors P1 and P2 are formed in the N well region NW, the NMOS transistor N1 and the NMOS transistors Q21 and Q22 are formed in the P well region PW1, and the NMOS transistor N2 and the NMOS transistors Q11 and Q12 are formed in the P well region PW0. The P well region PW0 and the P well region PW1 are formed opposite to each other with the N well region NW interposed therebetween.

In the N well region NW, the PMOS transistor P1 is constituted by a polysilicon wiring PL1 provided across a P+ diffusion region FLP1 and the PMOS transistor P2 is constituted by a polysilicon wiring PL2 provided across a P+ diffusion region FLP2.

In the P well region PW0, the NMOS transistor N1 is constituted by the polysilicon wiring PL1 provided across an N+ diffusion region FLN1 and the NMOS transistors Q11 and Q12 are constituted by a polysilicon wiring PL10 provided across N+ diffusion regions FLQ11 and FLQ12. The polysilicon wiring PL1 is formed from the N well region NW to the P well region PW1 and is thus shared as gates of the NMOS transistor N1 and the PMOS transistor P1, and the polysilicon wiring PL10 is shared as gates of the NMOS transistors Q11 and Q12.

In the P well region PW1, the NMOS transistor N2 is constituted by the polysilicon wiring PL2 provided across an N+ diffusion region FLN2 and the NMOS transistors Q21 and Q22 are constituted by a polysilicon wiring PL20 provided across N+ diffusion regions FLQ21 and FLQ22, respectively. The polysilicon wiring PL2 is formed from the N well region NW to the P well region PW0 and is thus shared as gates of the NMOS transistor N2 and the PMOS transistor P2, and the polysilicon wiring PL20 is shared as gates of the NMOS transistors Q21 and Q22.

Moreover, the polysilicon wiring PL1 and the polysilicon wiring PL20 are formed on the same straight line, the polysilicon wiring PL2 and the polysilicon wiring PL10 are formed on the same straight line, the diffusion regions FLP1, FLN1, FLQ21 and FLQ22 are formed on the same straight line, and the diffusion regions FLP2, FLN2, FLQ11 and FLQ12 are formed on the same straight line.

As a result, the PMOS transistor P1 and the NMOS transistors N1, Q21 and Q22 can be formed along the same straight line and the PMOS transistor P2 and the NMOS transistors N2, Q11 and Q12 can be formed along the same straight line. Consequently, a cell height HC2 of the 2-port memory cell MD can be set corresponding to two transistors in the same manner as the single port memory cell MS (HC2=HC1).

The P+ diffusion regions FLP1 and FLP2 are obtained by implanting and diffusing a P-type impurity and the N+ diffusion regions FLN1, FLN2, FLQ11, FLQ12, FLQ21 and FLQ22 are obtained by implanting and diffusing an N-type impurity. In the description of FIG. 5, each diffusion region has an upper region referred to as one of regions and a lower region referred to as the other region with respect to the polysilicon wiring in FIG. 5.

In the P well region PW0, a ground wiring LG1 provided on one of the regions of the N+ diffusion region FLN1 is electrically connected to one of the regions of the diffusion region FLN1 through a diffusion contact hole CH. The polysilicon wiring PL10 is electrically connected to a word line 1WL1 through a gate contact hole GC, and a bit line 1BL21 (the first layer aluminum wiring) provided on the other region of the diffusion region FLQ11 is electrically connected to the other region of the diffusion region FLQ11 through the diffusion contact hole CH, and the inverted bit line bar 1BL21 (the first layer aluminum wiring) provided on the other region of the diffusion region FLQ12 is electrically connected to the other region of the diffusion region FLQ12 through the diffusion contact hole CH.

Furthermore, the polysilicon wiring PL1 and one of the regions of the N+ diffusion region FLQ12 are electrically connected to each other through the shared contact SC formed from one of the regions of the N+ diffusion region FLQ11 to the polysilicon wiring PL1.

In the N well region NW, a power wiring LV1 provided on one of the regions of the diffusion region FLP1 is electrically connected to one of the regions of the diffusion region FLP1 through the diffusion contact hole CH, and the power wiring LV1 provided on the other region of the diffusion region FLP2 is electrically connected to the other region of the diffusion region FLP2 through the diffusion contact hole CH.

In the P well region PW1, the bit line 1BL22 (the first layer aluminum wiring) provided on one of the regions of the diffusion region FLQ21 is electrically connected to one of the regions of the diffusion region FLQ21 through the diffusion contact hole CH, and the inverted bit line bar 1BL22 (the first layer aluminum wiring) provided on one of the regions of the diffusion region FLQ22 is electrically connected to one of the regions of the diffusion region FLQ22 through the diffusion contact hole CH.

The word line 1WL2 (the first layer aluminum wiring) provided on the