Abstract: Disclosed is a semiconductor device using a gate electrode such as an SRAM, wherein the electrode pattern is a formed with fidelity to a reticle pattern through no complicated layout design. The gate electrode pattern is formed in an area smaller than that of a conventional semiconductor device. In a lithographic step using a reticle pattern provided with substantially linear gate electrode patterns, a projecting portion in which at least a part of a contact region is arranged is formed such that it is included in almost the center of a long side of a linear gate electrode pattern and a concave portion facing at least the entire length of the projecting portion is formed such that it is included in a long side opposite to the projecting portion between transistor regions of a reticle pattern.

Abstract: The present invention provides a semiconductor device including a first gate—gate electrode layer located in a first conductive layer and including gate electrodes of a first load transistor and a first driver transistor and a second gate—gate electrode layer located in the first conductive layer and including gate electrodes of a second load transistor and a second driver transistor. A first drain—drain connecting layer is located in a second conductive layer which is an upper layer of the first conductive layer and connects a drain of the first load transistor with a drain of the first driver transistor. A second drain—drain connecting layer is located in the second conductive layer and connects a drain of the second load transistor with a drain of the second driver transistor.

Abstract: Disclosed is a semiconductor integrated circuit device (e.g., an SRAM) having memory cells each of a flip-flop circuit constituted by a pair of drive MISFETs and a pair of load MISFETs, the MISFETs being cross-connected by a pair of local wiring lines, and having transfer MISFETs, wherein gate electrodes of all of the MISFETs are provided in a first level conductive layer, and the pair of local wiring lines are provided respectively in second and third level conductive layers. The local wiring lines can overlap and have a dielectric therebetween so as to form a capacitance element, to increase alpha particle soft error resistance. Moreover, by providing the pair of local wiring lines respectively in different levels, integration of the device can be increased.

Abstract: SRAM cell and method for fabricating the same, the SRAM cell including a first local interconnection connected between first terminals of the first access transistor, the first load transistor, and the first drive transistor, and gates of the second load transistor, and the second drive transistor electrically, and a second local interconnection connected between first terminals of the second access transistor, the second load transistor, and the second drive transistor, and gates of the first load transistor, and the first drive transistor electrically, thereby reducing an area of the SRAM cell.

Abstract: A full CMOS SRAM cell includes first and second active regions formed in a semiconductor substrate. A word line traverses first and second areas of the second active region, and first and second gate electrodes are arranged to be perpendicular to the word line. The first and second gate electrodes are parallel to each other and traverse the first and second active regions, respectively. A power line is electrically connected to a first common source region and is arranged parallel to the word line, the first common source region being the first active region between the first gate electrode and the second gate electrode. A ground line is electrically connected to a second common source region and is arranged parallel to the word line, the second common source region being the second active region between the first gate electrode and the second gate electrode. First and second bit lines are arranged to be perpendicular to the word line and parallel to each other.

Abstract: An SRAM device has STI regions separated by mesas and doped regions including source/drain regions, active areas, wordline conductors and contacts in a semiconductor substrate is made with a source region has 90° transitions in critical locations. Form a dielectric layer above the active areas. Form the wordline conductors above the active areas transverse to the active areas. The source and drain regions of a pass gate transistor are on the opposite sides of a wordline conductor. Form the sidewalls along the <100> crystal plane. Form the contacts extending down through to the dielectric layer to the mesas. Substrate stress is reduced because the large active area region formed in the substrate assures that the contacts are formed on the <100> surfaces of the mesas are in contact with the mesas formed on the substrate and that the <110> surfaces of the silicon of the mesas are shielded from the contacts.

Abstract: An SRAM memory cell includes two inverters connected in complement with each other. Each inverter includes one NMOS transistor and one PMOS transistor. The gate of the NMOS transistor in one inverter is connected to the drain of the NMOS transistor in the other inverter and this forms a first node. The drain of the NMOS transistor in one inverter is connected to the gate of the NMOS transistor in the other inverter and this forms a second node. The drain of an another PMOS transistor and the gate of still another PMOS transistor are connected to the first node. The drain of the still another PMOS transistor and the gate of the another PMOS transistor are connected to the second node. The gate capacitance and drain capacitance of these PMOS transistors is appended to the two nodes.

Abstract: Gate electrodes in an inverter section and a transfer section are formed only on element areas, and connected to each other by means of local interconnection layers. As a result, a memory cell of a very small size but a large capacity can be formed without considering a gate fringe or shortening phenomenon problem.

Abstract: A compact SRAM cell that incorporates refractory metal-silicon-nitrogen resistive elements as its pull-up transistors is described which includes a semi-conducting substrate, a pair of NMOS transfer devices formed vertically on the sidewalls of an etched substrate by a metal conductor providing electrical communication between an n+ region in the substrate and a bitline on top, a pair of pull-down nMOS devices on the substrate connected to ground interconnects, and a pair of vertical high-resistive elements formed of a refractory metal-silicon-nitrogen and function as a load for connecting to Vdd. The invention further describes a method for fabricating such compact SRAM cell.

Abstract: The present invention describes the use of two semiconductor layers, a thin film (TF) layer and a bulk Si wafer layer, to make high density and high speed merged logic and memory IC chips. The memory cells use three-dimensional (3D) SRAM structures. Two kinds of 3D logic cells are disclosed. 3D form of the differential cascode voltage switch (DCVS) architecture, and a 3D form of the DCVS with pass gate (DCVSPG) logic architecture. A high density “system on chip” architecture is described. The high density is achieved by locating large PMOS transistors in the TF Si layer, and the fast NMOS transistors in a bulk Si wafer layer. A single process sequence to simultaneously make the logic and memory circuits on the IC chip is also described.

Abstract: A MROM memory cell structure for storing multi level bit information is disclosed. First of all, a substrate is provided. The substrate has first and second trenches therein, wherein the first trench is deeper than second trench. A conformnal dielectric layer formed on sidewall and bottom of the first and second trenches. A conductive layer filled in the first and second trenches and on the substrate. A first doped region is formed under the first trench. A second doped region is formed under the second trench. A third doped region is formed in surface of the substrate and between the first and second trenches.

Abstract: An SRAM memory cell device comprises wordline and bitline inputs for enabling read/write access to memory cell contents, and, a diffusion region for maintaining a voltage to preserve memory cell content when the cell is not being accessed. The device further comprises a transistor device having a gate input for receiving a wordline voltage to turn off the transistor device when not performing memory cell read/write access; and, a gate oxide layer formed under the transistor device gate exhibiting resistance property for leaking current therethrough when the wordline voltage is applied to the gate input and the transistor device is off. The diffusion region receives voltage derived from the wordline voltage applied to said gate input to enable retention of said memory cell content in the absence of applied bitline voltage to thereby reduce power consumption.

Abstract: A semiconductor memory cell and forming method thereof utilizes a vertical select transistor to eliminate the problem of a large cell surface area in memory cells of the related art utilizing phase changes. A memory cell with a smaller surface area than the DRAM device of the related art is achieved by the present invention. Besides low power consumption during read operation, the invention also provides phase change memory having low power consumption even during write operation. Phase change memory also has stable read-out operation.

Abstract: A Scalable Two-Transistor Memory (STTM) cell array having a 4F2 unit cell area, where F is the minimum feature size. The data lines and the bit lines alternate and are adjacent to each other along the Y-axis direction, and the word lines are laid out along the X-axis direction. Each STTM cell consists of a floating gate MOS sensing transistor at the surface of a semiconductor substrate, with a vertical double sidewall gate multiple tunnel junction barrier programming MOS transistor on top of the sensing transistor. A data line connects all source regions of the programming transistors and a bit line connects all the source/drain regions of the sensing transistors in a column direction. A word line connects all double sidewall gate regions of programming transistors in a row direction. This invention also deals with a column addressing circuit as well as the driving method for the circuit.

Abstract: A memory cell of an SRAM has a full CMOS cell structure having successively aligned three wells of different conductivity types, and includes first and second contact holes extending from positions on first and second gates to positions above an impurity region of a predetermined MOS transistor, and formed in a self-aligned fashion with respect to the first and second gates, and first and second local interconnections formed in the contact holes, respectively.

Abstract: A dual-polycide semiconductor structure and method for forming the same having reduced dopant cross-diffusion. A conductive layer is formed over a polysilicon layer having a first region doped with a first dopant and a second region adjoining the first region at an interface doped with a second dopant. A region of discontinuity is then formed in the conductive layer located away from the interface. The conductive layer formed over the polysilicon gate overlaps the interface to provide electrical continuity between the first and second regions of the polysilicon gate, but also includes a region of discontinuity to reduce dopant cross-diffusion.

Abstract: According to one aspect of the present invention, a semiconductor memory device has: a semiconductor layer formed on an insulating film; and a memory cell array including a matrix arrangement of a plurality of memory cells each made up of first and second transistors connected in series, one side of each memory cell being connected to a bit line and the other side of each memory cell being supplied with a reference potential, and according to another aspect of the present invention, a semiconductor memory device manufacturing method includes: forming an oxide layer and a silicon active layer on a semiconductor substrate; forming an element isolation region for separating said silicon active layer into discrete element-forming regions to be substantially flush with said silicon active layer; forming gate electrode of paired two transistors by depositing a gate electrode material on said silicon active layer and patterning it; injecting predetermined ions into a region for forming a diffusion layer in, using sa

Abstract: An integrated static random access memory device includes four transistors and two resistors defining a memory cell. The four transistors are in a semiconductor substrate and are mutually interconnected by a local interconnect layer. The local interconnect layer is under a first metal level and a portion of the local interconnect layer defines above the substrate a base metal level. The two resistors extend in contact with a portion of the local interconnect layer between the base metal level and the first metal level.

Abstract: An SRAM of the present invention comprises a plurality of memory cells, which are formed over a plurality of wells, which store data and which do not have a well contact region for fixing the potential of the wells, and a plurality of well contact cells for fixing the potential of the wells that are formed over the plurality of wells so as to adjoin the memory cells, wherein the areas of the memory cells and the areas of the well contact cells are equal.

Abstract: The present invention provides a device design and method for forming the same that results in Fin Field Effect Transistors having different gains without negatively impacting device density. The present invention forms relatively low gain FinFET transistors in a low carrier mobility plane and relatively high gain FinFET transistors in a high carrier mobility plane. Thus formed, the FinFETs formed in the high mobility plane have a relatively higher gain than the FinFETs formed in the low mobility plane. The embodiments are of particular application to the design and fabrication of a Static Random Access Memory (SRAM) cell. In this application, the bodies of the n-type FinFETs used as transfer devices are formed along the {110} plane. The bodies of the n-type FinFETs and p-type FinFETs used as the storage latch are formed along the {100}.

Abstract: The invention provides SRAMs that can reduce memory cells in size and correct light proximity effect. Gate electrode layers in a first layer, drain—drain connection layers in a second layer, and drain-gate connection layers in a third layer define conduction layers of a flip-flop. The drain-gate connection layer has an extension section extending in a direction toward the drain-gate connection layer. The drain-gate connection layer 41b has an extension section extending in a direction toward the drain-gate connection layer.

Abstract: A semiconductor memory device includes a trench type SRAM(Static Random Access Memory) cell having a higher integration than a stack type SRAM. The SRAM cell memory device is provided with a trench formed in a semiconductor substrate and having four side walls therein, wherein a source region and a drain region of each of first and second drive transistors are formed in two of the four side walls. A pair of active layers respectively having a source region and a drain region of a first load transistor and a second load transistor, respectively, are formed on the substrate adjacent to the side walls. A gate electrode common to the first drive transistor and the first load transistor is formed on a gate oxide film. A gate electrode of an access transistor is vertically formed in a direction vertical to the semiconductor substrate instead of being formed on the substrate for thereby decreasing an area to be occupied by the transistor.

Abstract: The invention provides SRAMs that can reduce memory cells in size and correct light proximity effect. Gate electrode layers in a first layer, drain—drain connection layers in a second layer, and drain-gate connection layers in a third layer define conduction layers of a flip-flop. Driver transistors of one memory cell do not commonly share the n+ type source region with driver transistors of another memory cell.

Abstract: A semiconductor memory device (SRAM) comprises memory cells, each of which includes two load transistors, two driver transistors and two transfer transistors. The SRAM cell includes a semiconductor substrate in which the transistors are formed, a first interlayer dielectric formed on the semiconductor substrate, first contact portions formed in the first interlayer dielectric and first wiring layers (node wiring layers and pad layers) formed on the first interlayer dielectric. The first contact portions and the first wiring layers include metal layers made of refractory metal and a refractory metal nitride layers. This semiconductor memory device of the present invention is capable of enhancing an integration degree of wiring layers and achieving a microfabrication.

Abstract: A two transistor programmable logic cell 100 is used in programmable logic devices. The cell 100 has a backgate 3 that holds charge to program one of the two transistors into a logic 1 or a logic 0 state. Programmable logic devices are scalable to fine geometries and high densities and may be programmed to perform multiple logic functions on the same substrate.

Abstract: A dual-polycide semiconductor structure and method for forming the same having reduced dopant cross-diffusion. A conductive layer is formed over a polysilicon layer having a first region doped with a first dopant and a second region adjoining the first region at an interface doped with a second dopant. A region of discontinuity is then formed in the conductive layer located away from the interface. The conductive layer formed over the polysilicon gate overlaps the interface to provide electrical continuity between the first and second regions of the polysilicon gate, but also includes a region of discontinuity to reduce dopant cross-diffusion.

Abstract: A semiconductor memory device may be formed from a pair of transfer MOS transistors 1, 2 controlled by a word line 11 and a pair of data retaining flip-flop circuit formed from serially connected load elements 5, 6 and drive MOS transistors 3, 4. In the semiconductor memory device, the transfer MOS transistors 1, 2 have a threshold voltage greater than a threshold voltage of the drive MOS transistors 3, 4. The memory device may display an improved &bgr; ratio, and reduce the size of the drive MOS transistors to thereby reduce the cell area.

Abstract: A Scalable Two-Transistor Memory (STTM) cell array having a 4F2 unit cell area, where F is the minimum feature size. The data lines and the bit lines alternate and are adjacent to each other along the Y-axis direction, and the word lines are laid out along the X-axis direction. Each STTM cell consists of a floating gate MOS sensing transistor at the surface of a semiconductor substrate, with a vertical double sidewall gate multiple tunnel junction barrier programming MOS transistor on top of the sensing transistor. A data line connects all source regions of the programming transistors and a bit line connects all the source/drain regions of the sensing transistors in a column direction. A word line connects all double sidewall gate regions of programming transistors in a row direction. This invention also deals with a column addressing circuit as well as the driving method for the circuit.

Abstract: Provided are a semiconductor memory device and a method of manufacturing the same, in which landing pad layers in correspondence to contacts connecting to a power supply voltage line and a bit line can be easily formed in the same layer as node wiring, thereby simplifying the manufacturing process. Resist patterns having a roughly C shape i.e., patterns of two sets of node wiring are formed not in the same direction in all memory cells but in a different direction from each other between neighboring cells in up-down and right-left directions. Furthermore, out of four sides of each memory cell resist patterns i.e., patterns of the landing pad layer are formed on the two sides opposite the two sides close to the resist patterns for the two sets of node wiring. Accordingly, the landing pad layers can be formed in the same layer as the node wiring, the landing pad layers corresponding to contacts connecting to a grounded line, power supply voltage line and bit line in the upper layer from the node wiring.

Abstract: N well contact area 13 is integrally formed with second diffused area 12 within the upper parts of a N well and a P well, and P well contact area 14 is integrally formed with first diffused area 11 in the upper parts of the P well and the N well.

Abstract: With a P well region being divided, NMOS transistors N1 and N3 are formed in the first P well region, and NMOS transistors N2 and N4 in the second P well region. Alternatively, with a N well region being divided, PMOS transistor P1 is formed in the first N well region, and PMOS transistor P2 in the second N well region.

Abstract: A full CMOS SRAM cell includes first and second active regions formed in a semiconductor substrate. A word line traverses first and second areas of the second active region, and first and second gate electrodes are arranged to be perpendicular to the word line. The first and second gate electrodes are parallel to each other and traverse the first and second active regions, respectively. A power line is electrically connected to a first common source region and is arranged parallel to the word line, the first common source region being the first active region between the first gate electrode and the second gate electrode. A ground line is electrically connected to a second common source region and is arranged parallel to the word line, the second common source region being the second active region between the first gate electrode and the second gate electrode. First and second bit lines are arranged to be perpendicular to the word line and parallel to each other.

Abstract: A static random access memory comprising memory cells each composed of transfer MISFETs controlled by word lines and of a flip-flop circuit made of driver MISFETs and load MISFETs. The top of the load MISFETs is covered with supply voltage lines so that capacitor elements of a stacked structure are formed between the gate electrodes of the load MISFETs and the supply voltage lines.

Abstract: A static semiconductor memory device capable of preventing soft errors is provided. The static semiconductor memory device includes: a silicon substrate having a p-type well region; a storage node; an n-type-low-concentration impurity region and a high-concentration impurity region formed in the surface of p-type well region and connected to storage node; and a p-type impurity region formed to have contact with high-concentration impurity region.

Abstract: The invention provides SRAMs that can reduce memory cells in size and correct light proximity effect. Gate electrode layers in a first layer, drain-drain connection layers in a second layer, and drain-gate connection layers in a third layer define connection wirings of a flip-flop. A p+ type well contact region is provided for every two of the memory cells arranged in the Y-axis direction.

Abstract: In a semiconductor device, a first conductive layer (2) is located on a semiconductor substrate (14) through an insulating film (13a) and beneath a first insulating layer (13f). On the first insulating layer (13f) is formed a second conductive layer (8) followed by a second insulating layer (13g), either or both of which are very thin. A third conductive layer (6) is placed on the second insulating layer (13g). A connecting column (16) extends from the third conducting layer (6) through and forming an end contact with the second conductive layer (8) to the first conducting layer (2) and the substrate (14), with a greater portion of the column resting upon the substrate (14). The third conductive layer (6) forms the gate electrode (6b) of a top gate type TFT.

Abstract: A semiconductor substrate has a main surface, a well, a plurality of memory cells, a first memory cell region, a second memory cell region, a border region, a well contact region, a first dummy element, a second dummy element, a first transistor and a second transistor. The first and second memory cell regions are located over the well. The memory cells are formed in the first and second memory cell regions. The border region is located over the well on a border between the first memory cell region and the second memory cell region. The well contact region is formed in the well in the border region and is electrically connected to a wiring layer for fixing the voltage of the well. The first and second dummy elements are formed in the border. The first transistor, that is a component of the memory cell, is formed in the first memory cell region and is located adjacent to the first dummy element.

Abstract: An SRAM of the present invention comprises a plurality of memory cells, which are formed over a plurality of wells, which store data and which do not have a well contact region for fixing the potential of the wells, and a plurality of well contact cells for fixing the potential of the wells that are formed over the plurality of wells so as to adjoin the memory cells, wherein the areas of the memory cells and the areas of the well contact cells are equal.

Abstract: In order to improve the performance of a semiconductor integrated circuit device wherein a capacitor provided between storage nodes of an SRAM and a device having an analog capacitor are formed on a single substrate, a plug is formed in a silicon oxide film on a pair of n channel type MISFETs in a memory cell forming area, and a local wiring LIc for connecting respective gate electrodes and drains of the pair of n channel type MISFETs is formed over the silicon oxide film and the plug. Thereafter, a capacitive insulating film and an upper electrode are further formed over the local wiring LIc. According to the same process step as the local wiring, capacitive insulating film and upper electrode formed in the memory cell forming area, a local wiring LIc, a capacitive insulating film and an upper electrode are formed over a silicon oxide film in an analog capacitor forming area and a plug in the silicon oxide film.

Abstract: In a semiconductor device including a semiconductor substrate, a well formed on the semiconductor substrate, and a thick field insulating layer for surrounding an active area of the well, a contact structure is buried in a contact hole provided in the thick field insulating layer and connected to the well, so as to fix a voltage at the well.

Abstract: A compact SRAM cell that incorporates refractory metal-silicon-nitrogen resistive elements as its pull-up transistors is described which includes a semi-conducting substrate, a pair of NMOS transfer devices formed vertically on the sidewalls of an etched substrate by a metal conductor providing electrical communication between an n+ region in the substrate and a bitline on top, a pair of pull-down nMOS devices on the substrate connected to ground interconnects, and a pair of vertical high-resistive elements formed of a refractory metal-silicon-nitrogen and function as a load for connecting to Vdd. The invention further describes a method for fabricating such compact SRAM cell.

Abstract: A semiconductor device and a method of producing the semiconductor device, fabricated by forming a memory device and a logic device on a single semiconductor substrate, are provided. A side wall (9) and a silicide protection film (10) of a gate electrode (7e) are used instead of forming a silicide protection film in a logic device region (101), whereby the number of steps in forming a logic process consolidating device can be reduced. Further, high concentration impurity regions are formed using the silicide protection film (10) as a mask, whereby a degree of freedom of a condition of implanting ions becomes high.

Abstract: On a semiconductor substrate surface, a drain diffusion layer, which is in common to two transistors that make up a memory cell pair, is formed and source diffusion layers, for each of the transistors, respectively, are formed so as to sandwich the drain diffusion layer from both sides, a bit line is formed from a lower wiring layer and is connected to the drain diffusion layer, a source line is formed from the uppermost wiring layer, and the writing of information is performed by making a contact hole exist or non-existent immediately below the source line arranged from the uppermost wiring layer, in other words, by connection or non-connection of the source diffusion layer with the source line. By this arrangement, the TAT can be shortened and, since the capacitance of the bit line is not increased, high-speed operation with a short precharge time and discharge time for the bit line can be realized and the consumption power can be lessened.

Abstract: A full CMOS SRAM cell includes first and second active regions formed in a semiconductor substrate. A word line traverses first and second areas of the second active region, and first and second gate electrodes are arranged to be perpendicular to the word line. The first and second gate electrodes are parallel to each other and traverse the first and second active regions, respectively. A power line is electrically connected to a first common source region and is arranged parallel to the word line, the first common source region being the first active region between the first gate electrode and the second gate electrode. A ground line is electrically connected to a second common source region and is arranged parallel to the word line, the second common source region being the second active region between the first gate electrode and the second gate electrode. First and second bit lines are arranged to be perpendicular to the word line and parallel to each other.

Abstract: An SRAM memory cell relating to the present invention includes: first and second access MOS transistors; first and second driver MOS transistors; and first and second load MOS transistors. An insulating layer is formed on first and second gates respectively forming gates of the first and second driver MOS transistors and gates of the first and second load MOS transistors. On the insulating layer, formed are first and second conductive layers for forming capacitances between the first and second gates and the conductive layers. Furthermore, formed are a first local interconnect connecting the first gate and the second conductive layer therebetween and a second local interconnect connecting the second gate and the first conductive layer therebetween.

Abstract: A semiconductor memory device which can suppress the occurrence of corner rounding through the resist patterning process to achieve a reduction in cell size and higher integration. A relationship between a channel width DT.W of the drive transistor, a channel length DT.L of the drive transistor, a channel width WT.W of the word transistor and a channel length WT.L of the word transistor is given by: (DT.W/WT.W)/(WT.L/DT.L)<1.2. The channel width DT.W of the drive transistor is equal to the channel width WT.W of the word transistor, to reduce steps in the patterns of p-type active regions. The channel length WT.L of the word transistor is larger than the channel length DT.L of the drive transistor, that is, (WT.L/DT.L)>1.

Abstract: The present invention provides a circuit, in which a device typified by a PLED element is built into a flip-flop. In this case, a storage node of the device is low leakage. According to the present invention, it is possible to realize a SRAM that has nonvolatility while achieving high-speed operation. It is also possible to realize a flip-flop having the same characteristics. An example of a typical mode of the present invention is a storage circuit characterized by the following: a storage element is a device incorporating: a first path for a carrier; a first mode for storing a charge that generates an electric field where conductivity of the first path is changed; and a barrier structure through which a second carrier moves in response to given voltage so that the second carrier is stored in the first node; and the storage circuit includes a second node, to which information stored in the first node is outputted steadily in a state in which power is supplied.

Abstract: A core section complementary transistor and a memory cell section complementary transistor are formed on a semiconductor substrate of a first conductivity type. The core section complementary transistor has a first well of a second conductivity type provided in the semiconductor substrate, a first core section MOS transistor provided on the first well of the second conductivity type, a second core section MOS transistor provided on the semiconductor substrate a device separation film which separates the first core section MOS transistor and the second core section MOS transistor from each other, and a well of the first conductivity type provided under a part of the device separation film which is closer to the second core section MOS transistor. The first core section MOS transistor has source-drain regions of the first conductivity type. The second core section MOS transistor has source-drain regions of the second conductivity type.

Abstract: A semiconductor body having a pair of vertical, double-gated CMOS transistors. An insulating layer extending horizontally beneath the surface of the semiconductor body such insulating layer being disposed beneath the pair of transistors. The transistors, together with additional such transistors, are arranged to form a Synchronous Dynamic Random Access Memory (SRAM) array. The array includes a plurality of SRAM cells arranged in rows and columns, each one of the cells having a WORDLINE connected to a WORLDINE CONTACT. The WORDLINE CONTACT is common to four contiguous one of the cells. One of the cells having a plurality of electrically interconnected MOS transistors arranged to provide an SRAM circuit. Each one of the cells has a VDD CONTACT and a VSS CONTACT. One of such CONTACTs is disposed centrally within each one of the cells and the other one of the CONTACTs being common to four contiguous ones of the cells.

Abstract: A first level metal interconnection line in a layer below a third level metal interconnection line serving as a main word line MWL is used as a shunting interconnection line and electrically connected to a first level polysilicon interconnection line constituting a sub word line SWL at prescribed intervals. By applying a hierarchical word line structure and a word line shunting structure both, a word line is driven into a selected state at high speed without increasing an array occupancy area and manufacturing steps.