Abstract: A method for operating a memory device includes applying a sequence of bias arrangements across a selected metal-oxide memory element to change among resistance states. The sequence of bias arrangements includes a first set of one or more pulses to change the resistance state of the selected metal-oxide memory element from the first resistance state to a third resistance state, and a second set of one or more pulses to change the resistance state of the selected metal-oxide memory element from the third resistance state to the second resistance state.

Abstract: A method of forming a semiconductor memory device includes forming a tunnel insulating layer on a semiconductor substrate, and forming a silicon layer, including metal material, on the tunnel insulating layer. Accordingly, an increase in the strain energy of the conductive layer may be prohibited and, therefore, the growth of grains constituting the conductive layer may be prevented. Furthermore, a threshold voltage distribution characteristic and electrical properties of a semiconductor memory device may be improved.

Abstract: A semiconductor device includes an active region defined by an isolation region formed in a cell area, buried gates disposed in the active region and the isolation region, conduction layers disposed on the active region and having the same heights as an surface of the isolation region, and a line type storage node contact connected with one of the conduction layers.

Abstract: A nonvolatile memory device and a method of fabricating the same is provided to prevent charges stored in a charge trap layer from moving to neighboring memory cells. The method of fabricating a nonvolatile memory device, includes forming a first dielectric layer on a semiconductor substrate in which active regions are defined by isolation layers, forming a charge trap layer on the first dielectric layer, removing the first dielectric layer and the charge trap layer over the isolation layers, forming a second dielectric layer on the isolation layers including the charge trap layer, and forming a conductive layer on the second dielectric layer.

Abstract: A non-volatile memory device includes a plurality of pillars, where each of the plurality of pillars contains a non-volatile memory cell containing a steering element and a storage element and at least one of a top corner or a bottom corner of each of the plurality of pillars is rounded. A method of making non-volatile memory device includes forming a stack of device layers, and patterning the stack to form a plurality of pillars, where each of the plurality of pillars contains a non-volatile memory cell that contains a steering element and a storage element, and where at least one of top corner or bottom corner of each of the plurality of pillars is rounded.

Abstract: A method for forming a floating gate semiconductor device such as an electrically erasable programmable read only memory is provided. The device includes a silicon substrate having an electrically isolated active area. A gate oxide, as well as other components of a FET (e.g., source, drain) are formed in the active area. A self aligned floating gate is formed by depositing a conductive layer (e.g., polysilicon) into the recess and over the gate oxide. The conductive layer is then chemically mechanically planarized to an endpoint of the isolation layer so that all of the conductive layer except material in the recess and on the gate oxide is removed. Following formation of the floating gate an insulating layer is formed on the floating gate and a control gate is formed on the insulating layer.

Abstract: A method of fabricating a vertical transistor in a semiconductor device improves integration of the semiconductor device according to a design rule. After a semiconductor substrate is etched to form a buried bit line, a gate electrode pattern that surrounds a cylindrical channel region pattern of the vertical transistor is formed, thereby preventing damage to the gate electrode pattern due to an etching process. The gate electrode pattern surrounds the channel region pattern where a width is narrower than second source and drain regions. The second source and drain regions are then deposited over the channel region pattern and the gate electrode pattern. As a result, a neck-shaped channel region does not collapse due to the weight of the second source and drain regions.

Abstract: Some embodiments include methods of forming memory arrays. A stack of semiconductor material plates may be patterned to subdivide the plates into pieces. Electrically conductive tiers may be formed along sidewall edges of the pieces. The pieces may then be patterned into an array of wires, with the array having vertical columns and horizontal rows. Individual wires may have first ends joining to the electrically conductive tiers, may have second ends in opposing relation to the first ends, and may have intermediate regions between the first and second ends. Gate material may be formed along the intermediate regions. Memory cell structures may be formed at the second ends of the wires. A plurality of vertically-extending electrical interconnects may be connected to the wires through the memory cell structures, with individual vertically-extending electrical interconnects being along individual columns of the array. Some embodiments include memory arrays incorporated into integrated circuitry.

Abstract: Disclosed is a method of manufacturing a semiconductor device. The method includes forming an oxide-nitride-oxide (ONO) layer over a semiconductor substrate, and forming a recess over the semiconductor substrate by etching the ONO layer, forming a vertical structure pattern being higher than the ONO layer over the recess, sequentially forming a spacer oxide film and a first gate poly over the side wall of the vertical structure pattern, and forming a nitride film spacer at a partial region of the side wall of the first gate poly, removing the nitride film spacer, and forming a second gate poly in a spacer shape over the side wall of the first gate poly, and forming a first split gate and a second split gate, symmetrically divided from each other, by removing the vertical structure pattern.

Abstract: A semiconductor device is provided with a conductor wire and a fuse wire formed in an insulating film over a semiconductor substrate, a first under-pad-wire insulating film formed above the insulating film, a second under-pad-wire insulating film formed on the first under-pad-wire insulating film, a pad wire formed in an area above the conductive wire, in the first and second under-pad-wire insulating films and an opening formed by leaving a part of the first under-pad-wire insulating film in an area above the fuse wire, in the first and second under-pad-wire insulating films, wherein the second under-pad-wire insulating film comprises an element different from that of the first under-pad-wire insulating film.

Abstract: A stylus, an integrated circuit (IC) and method of forming the IC. The stylus extends upward from its apex and has a substantially circular cross section that decreases in diameter upward from the apex. The stylus is formed in a mold that may be formed in an orifice in a dielectric layer between wiring layers. The mold may include multiple concentric layers. For a more pronounced, non-linear stylus taper, each layer may be thinner than its next adjacent outer concentric layer.

Abstract: To provide a technology capable of improving the property of an MRAM in a semiconductor device containing the MRAM. A plasma treatment is performed on the surface of an interlayer insulating film for which a wiring and a digit line are formed. Firstly, a semiconductor substrate is carried in a chamber, and a mixed gas that includes molecules containing nitrogen (ammonia gas) and inert molecules not containing nitrogen (hydrogen gas, helium, argon) is introduced into the chamber. On this occasion, the plasma treatment is performed by introducing the mixed gas under such a condition that the flow rate of the inert molecules not containing nitrogen is larger than that of the molecules containing nitrogen, and the mixed gas is turned into a plasma.

Abstract: The invention provides an advanced metallization technique for fabricating a memory cell array on a substrate. The array is fabricated by forming discrete and self-aligned vias in a first layer disposed over the array to form contacts to each of the source and drain junction in the array. Further, self-aligned local area slotted vias are formed in a second layer that is disposed over the first layer to form local area interconnects that electrically shunt all of the source contacts/junctions. Further, discrete self-aligned drain extensions are formed over each of the formed drain contacts to electrically connect the junctions, and source contacts to the extensions. The formed vias, extensions, and slotted local area vias are simultaneously plugged and filled with a conductive material to form the memory cell array.

Abstract: A memory includes a first tunneling field effect transistor including a first drain and a first source, the first drain coupled to a first resistive memory element. The memory includes a second tunneling field effect transistor including a second drain and sharing the first source, the second drain coupled to a second resistive memory element. The memory includes a first region coupled to the first source for providing a source node.

Abstract: The use of an ammonium hydroxide spike to a hot tetra methyl ammonium hydroxide (TMAH) solution to form an insitu poly oxide decapping step in a polysilicon (poly) etch process, results in a single step rapid poly etch process having uniform etch initiation and a high etch selectivity, that may be used in manufacturing a variety of electronic devices such as integrated circuits (ICs) and micro electro-mechanical (MEM) devices. The etching solution is formed by adding 35% ammonium hydroxide solution to a hot 12.5% TMAH solution at about 70° C. at a rate of 1% by volume, every hour. Such an etch solution and method provides a simple, inexpensive, single step self initiating poly etch that has etch stop ratios of over 200 to 1 over underlying insulator layers and TiN layers.

Abstract: A method of fabricating a semiconductor memory device includes alternately and repeatedly stacking sacrificial layers and insulating layers on a substrate, forming an active pattern penetrating the sacrificial layers and the insulating layers, continuously patterning the insulating layers and the sacrificial layers to form a trench, removing the sacrificial layers exposed in the trench to form recess regions exposing a sidewall of the active pattern, forming an information storage layer on the substrate, forming a gate conductive layer on the information storage layer, such that the gate conductive layer fills the recess regions and defines an empty region in the trench, the empty region being surrounded by the gate conductive layer, and performing an isotropic etch process with respect to the gate conductive layer to form gate electrodes in the recess regions, such that the gate electrodes are separated from each other.

Abstract: Circuits and methods for providing a dual gate oxide (DGO) embedded SRAM with additional logic portions, where the logic and the embedded SRAM have NMOS transistors having a common gate dielectric thickness but have different lightly doped drain (LDD) implantations formed using different LDD masks to provide optimum transistor operation. In an embodiment, a first embedded SRAM is a single port device and a second embedded SRAM is a dual port device having a separate read port. In certain embodiments, the second SRAM includes NMOS transistors having LDD implants formed using the logic portion LDD mask. Transistors formed with the logic portion LDD mask are faster and have lower Vt than transistors formed using a SRAM LDD mask. Dual core devices having multiple embedded SRAM arrays are disclosed. Methods for making the embedded SRAM are also disclosed.

Abstract: One or more embodiments relate to a method of forming a semiconductor device, including: providing a substrate; forming a gate stack over the substrate, the gate stack including a control gate over a charge storage layer; forming a conductive layer over the gate stack; etching the conductive layer to remove a portion of the conductive layer; and forming a select gate, the forming the select gate comprising etching a remaining portion of the conductive layer.

Abstract: A non-volatile memory device includes a first electrode, a diode steering element, at least three resistivity switching storage elements, and a second electrode. The diode steering element electrically contacts the first electrode and the at least three resistivity switching storage elements. The second electrode electrically contacts only one of the at least three resistivity switching storage elements.

Abstract: In one aspect of the present invention, a programmable element, may include a semiconductor substrate, source/drain layers formed apart from each other in the upper surface of the semiconductor substrate, a gate insulating film including a charge-trapping film containing Hf and formed on a portion between the source/drain layers of the semiconductor substrate, and a gate electrode formed on the gate insulating film with a program voltage applied to the gate electrode.

Abstract: A method of improving high aspect ratio etching by reverse masking to provide a more uniform mask height between the array and periphery is presented. A layer of amorphous carbon is deposited over a substrate. An inorganic hard mask is deposited on the amorphous carbon followed by a layer of photodefinable material which is deposited over the array portion of the substrate. The photodefinable material is removed along with the inorganic hard mask overlaying the periphery. A portion of the amorphous carbon layer is etched in the exposed periphery. The inorganic hard mask is removed and normal high aspect ratio etching continues. The amount of amorphous carbon layer remaining in the periphery results in a more uniform mask height between the array and periphery at the end of high aspect ratio etching. The more uniform mask height mitigates twisting at the edge of the array.

Abstract: In a method of fabricating a semiconductor device having vertical channels and a method of patterning a gate electrode of such semiconductor device, an initial conductive layer is removed by multiple etching processes.

Abstract: A method of manufacturing a semiconductor device, comprising: forming, on a semiconductor substrate a gate insulating film for a high-voltage transistor of a peripheral circuit; forming on the gate insulating film a gate electrode for the high-voltage transistor; removing the gate insulating film positioned on the semiconductor substrate on both side portions of the gate electrode; forming an impurity diffusion region in a surface of the semiconductor substrate; depositing a first silicon oxide film to extend over surfaces of the gate electrode and the impurity diffusion region; etching the first silicon oxide film to form a spacer such that the spacer is formed on a side wall portion of the gate electrode and also extends over the surface of the semiconductor substrate; and forming a silicon nitride film on a surface of the spacer.

Abstract: A system for terminating a homogenous cell array is disclosed. A preferred embodiment comprises a plurality of homogenous cells arranged in rows and columns to form the homogenous cell array, wherein a first homogenous cell of each column is electrically differently connected than a rest of the homogenous cells of the column.

Abstract: A semiconductor device manufacturing method includes: alternately stacking a plurality of insulating layers and electrode layers; forming a hole penetrating through a multilayer body of the insulating layers and the electrode layers; forming a conductive film on an inner wall of the hole; anisotropically etching the conductive film to selectively leave the conductive film on a sidewall of the hole; altering the conductive film into an insulator by heat treatment; and removing the insulator covering the electrode layers to expose the electrode layers into the hole.

Abstract: A semiconductor device has two transistors of different structure from each other. One of transistors is P-type and the other is N-type. One of the transistors includes a gate structure in which a polysilicon layer contacts a gate insulation film while the other transistor includes a gate structure in which a metal layer contacts a gate insulation film.

Abstract: Embodiments of the present invention provide a method that includes depositing a first electrode film on one or more wordline structures, depositing a phase change material (PCM) film on the first electrode film, depositing a second electrode film on the PCM film, depositing a third electrode film on the second electrode film, depositing an access device film on the third electrode film, and depositing a fourth electrode film on the access device film to form a stack of films, wherein the stack of films comprises the first electrode film, the PCM film, the second electrode film, the third electrode film, the access device film, and the fourth electrode film. Other embodiments may be described and/or claimed.

Abstract: To improve performance of a semiconductor device. Over a semiconductor substrate, a plurality of p-channel type MISFETs for logic, a plurality of n-channel type MISFETs for logic, a plurality of p-channel type MISFETs for memory, and a plurality of n-channel type MISFETs for memory are mixedly mounted. At least a part of the p-channel type MISFETs for logic have each a source/drain region constituted by silicon-germanium, and all the n-channel type MISFETs for logic have each a source/drain region constituted by silicon. All the p-channel type MISFETs for memory have each a source/drain region constituted by silicon, and all the n-channel type MISFETs for memory have each a source/drain region constituted by silicon.

Abstract: A phase-change memory element is provided. The phase-change memory element includes: a first electrode formed on a substrate; a first dielectric layer, with an opening, formed on the first electrode, wherein the opening exposes a top surface of the first electrode; a pillar structure formed directly on the first electrode within the opening; an inner phase-change material layer surrounding the pillar structure, directly contacting the first electrode; a second dielectric layer surrounding the inner phase-change material layer; an outer phase-change material layer surrounding the second dielectric layer; a phase-change material collar formed between the second dielectric layer and the first electrode, connecting the inner phase-change material layer with the outer phase-change material layer; and a second electrode formed directly on the pillar structure, directly contacting the top surface of the inner phase-change material layer.

Abstract: Nanowire-based devices are provided. In one aspect, a SRAM cell includes at least one pair of pass gates and at least one pair of inverters formed adjacent to one another on a wafer. Each pass gate includes one or more device layers each having a source region, a drain region and a plurality of nanowire channels connecting the source region and the drain region and a gate common to each of the pass gate device layers surrounding the nanowire channels. Each inverter includes a plurality of device layers each having a source region, a drain region and a plurality of nanowire channels connecting the source region and the drain region and a gate common to each of the inverter device layers surrounding the nanowire channels.

Abstract: A method for manufacturing a semiconductor device includes the consecutive steps of selectively implanting first-conductivity-type impurities into a silicon substrate in a memory cell array area to form first source/drain regions, heat treating to diffuse the impurities in the first source/din regions; selectively implanting impurities into the silicon substrate in a peripheral circuit area to form second source/drain regions in the peripheral circuit area.

Abstract: A method of manufacturing a nonvolatile storage device having memory cell arrays according to an embodiment of the present invention includes forming, in a memory cell array forming region above a processed film, first columnar members arrayed at substantially equal intervals in the first direction and the second direction, forming, concerning at least arrays as a part of arrays of the first columnar members in the first direction, second columnar members long in section having major axes longer than sections of the first columnar members outside of the memory cell array forming region such that the major axes are set in the first direction and the second columnar members continue to ends of the arrays, and forming, in the same manner as above, third columnar members, which continue to arrays of the first columnar members in the second direction.

Abstract: An array with cells that have adjacent similar structures that are displaced from each other across a common cell border in a direction that is not perpendicular to the cell border thus avoiding an across cell border design rule violation between the adjacent similar structures. A method of forming reduced area memory arrays by displacing adjacent similar structures along a common cell border. A method of building arrays using conventional array building software by forming unit pairs with cells that are not identical and are not mirror images or rotated versions of each other.

Abstract: An array with cells that have adjacent similar structures that are displaced from each other across a common cell border in a direction that is not perpendicular to the cell border thus avoiding an across cell border design rule violation between the adjacent similar structures. A method of forming reduced area memory arrays by displacing adjacent similar structures along a common cell border. A method of building arrays using conventional array building software by forming unit pairs with cells that are not identical and are not mirror images or rotated versions of each other.

Abstract: The use of atomic layer deposition (ALD) to form a nanolaminate dielectric of zirconium oxide (ZrO2), hafnium oxide (HfO2) and tin oxide (SnO2) acting as a single dielectric layer with a formula of Zrx Hfy Sn1-x-y O2, and a method of fabricating such a dielectric layer is described that produces a reliable structure with a high dielectric constant (high k). The dielectric structure is formed by depositing zirconium oxide by atomic layer deposition onto a substrate surface using precursor chemicals, followed by depositing hafnium oxide onto the substrate using precursor chemicals, followed by depositing tin oxide onto the substrate using precursor chemicals, and repeating to form the thin laminate structure. Such a dielectric may be used as a gate insulator, a capacitor dielectric, or as a tunnel insulator in non-volatile memories, because the high dielectric constant (high k) provides the functionality of a much thinner silicon dioxide film.

Abstract: The present invention provides a technology capable of reducing an area occupied by a nonvolatile memory while improving the reliability of the nonvolatile memory. In a semiconductor device, the structure of a code flash memory cell is differentiated from that of a data flash memory cell. More specifically, in the code flash memory cell, a memory gate electrode is formed only over the side surface on one side of a control gate electrode to improve a reading speed. In the data flash memory cell, on the other hand, a memory gate electrode is formed over the side surfaces on both sides of a control gate electrode. By using a multivalued memory cell instead of a binary memory cell, the resulting data flash memory cell can have improved reliability while preventing deterioration of retention properties and reduce its area.

Abstract: A nonvolatile memory integrated circuit has a semiconductor substrate and a nonvolatile memory device on the semiconductor substrate. The device has a transistor and a capacitor on the semiconductor substrate, and a shared floating gate connecting the gate regions of the transistor and the capacitor. The transistor has at least a doping region defining the source and drain regions, as well as three other doping regions overlapping the source and drain regions. Also disclosed are a nonvolatile memory circuit with multiple such nonvolatile memory device, and methods for making the nonvolatile memory circuit with one or more such nonvolatile memory devices.

Abstract: Methods for fabricating dual bit memory devices are provided. In an exemplary embodiment of the invention, a method for fabricating a dual bit memory device comprises forming a charge trapping layer overlying a substrate and etching an isolation opening through the charge trapping layer. An oxide layer is formed overlying the charge trapping layer and within the isolation opening. A control gate is fabricated overlying the isolation opening and portions of the charge trapping layer adjacent to the isolation opening. The oxide layer and the charge trapping layer are etched using the control gate as an etch mask and impurity dopants are implanted into the substrate using the control gate as an implantation mask.

Abstract: The present invention relates to a highly integrated semiconductor device in which a capacitor is formed between adjacent gate patterns by using a nanotube process. A semiconductor memory device according to an example embodiment of the present invention includes a capacitor formed on a first side of a source/drain region positioned between gate patterns of adjacent cell transistors; a plate layer connected to an upper portion of the capacitor, the plate layer being formed in a direction intersecting the gate pattern; and a bit line connected to a second side of the source/drain region of the cell transistor, the bit line being formed in the direction intersecting the gate pattern.

Abstract: A semiconductor device includes a device isolation layer in a semiconductor substrate, an active region defined by the device isolation layer, the active region including a main surface and a recess region including a bottom surface that is lower than the main surface, and a gate electrode formed over the recess region, wherein a top surface of the device isolation layer adjacent to the recess region is lower than the bottom surface of the recess region.

Abstract: A nonvolatile memory including a plurality of memory transistors in series, wherein source/drain and channel regions therebetween are of a first type and a select transistor, at each end of the plurality of memory transistors in series, wherein channels regions of each of the select transistors is of the first type. The first type may be n-type or p-type. The nonvolatile memory may further include a first dummy select transistor at one end of the plurality of memory transistors in series between one of the select transistors and the plurality of memory transistors in series and a second dummy select transistor at the other end of the plurality of memory transistors in series between the other select transistor and the plurality of memory transistors in series.

Abstract: A method includes forming an electrical insulator material over an integrated circuit having a metal-containing conductive interconnect and activating a dopant in a semiconductor material of a substrate to provide a doped region. The doped region provides a junction of opposite conductivity types. After activating the dopant, the substrate is bonded to the insulator material and at least some of the substrate is removed where bonded to the insulator material. After the removing, a memory cell is formed having a word line, an access diode, a state-changeable memory element containing chalcogenide phase change material, and a bit line all electrically connected in series, the access diode containing the junction as a p-n junction. A memory device includes an adhesion material over the insulator material and bonding the word line to the insulator material.

Abstract: A cell array includes a semiconductor substrate including an active region comprising a first region, a second region, and a transition region, the second region being separated from the first region by the transition region, wherein a top surface of the second region is at a different level than a top surface of the first region. The cell array also includes a plurality of word lines crossing over the first region. The cell array also includes a selection line crossing over the active region, wherein at least a portion of the selection line is located over the transition region.

Abstract: A method for forming a pattern in a semiconductor device includes forming an etch target layer comprising metal over a substrate. A hard mask pattern is formed over the etch target layer. The etch target layer is etched to form a pattern such that a line width of the etch target layer is smaller than a line width of the hard mask pattern.

Abstract: A non-volatile semiconductor storage device includes: a substrate; a control circuit layer provided on the substrate; a support layer provided on the control circuit layer; and a memory cell array layer provided on the support layer. The memory cell array layer includes: a first lamination part having first insulation layers and first conductive layers alternately laminated therein; and a second lamination part provided on either the top or bottom surface of the respective first lamination part and laminated so as to form a second conductive layer between second insulation layers. The control circuit layer includes at least any one of: a row decoder driving word lines provided in the memory cell array layer, and a sense amplifier sensing and amplifying a signal from bit lines provided in the memory cell array layer.

Abstract: An integrated circuit structure includes a dielectric layer having an upper portion and a lower portion. The dielectric layer is either an inter-layer dielectric (ILD) or an inter-metal dielectric (IMD). A phase change random access memory (PCRAM) cell includes a phase change strip, wherein the phase change strip is on the lower portion and has a top surface lower than a top surface of the dielectric layer, and a bottom surface higher than a bottom surface of the dielectric layer. A first conductive column is electrically connected to the phase change strip. The first conductive column extends from the top surface of the dielectric layer down into the dielectric layer. A second conductive column is in a peripheral region. The second conductive column extends from the top surface of the dielectric layer down into the dielectric layer. The first conductive column and the second conductive column have different heights.

Abstract: A semiconductor memory device according to one embodiment includes: a semiconductor substrate having an active region divided by an element isolation region; a plurality of stacked-gate type memory cell transistors connected in series on the active region; select transistors connected to both ends of the plurality of memory cell transistors on the active region; and a bit line contact connected to a drain region belonging to the select transistor in the active region, a vertical cross sectional shape of a lower portion of the bit line contact in a channel width direction of the plurality of memory cell transistors being in a skirt shape.

Abstract: A semiconductor device according to an embodiment includes: a semiconductor substrate; a resistance element of a first conductivity type formed in one region of the semiconductor substrate; a field effect transistor of a second conductivity type formed in another region of the semiconductor substrate; and a field effect transistor of the first conductivity type formed in still another region of the semiconductor substrate. The resistance element includes: an insulating film formed in an upper layer portion of the semiconductor substrate; and a well of the first conductivity type formed immediately below the insulating film, an impurity concentration at an arbitrary depth position in the well of the first conductivity is lower than an impurity concentration at the same depth position in a channel region of the field effect transistor of the second conductivity type.

Abstract: In a phase change memory, electric property of a diode used as a selection device is extremely important. However, since crystal grain boundaries are present in the film of a diode using polysilicon, it involves a problem that the off leak property varies greatly making it difficult to prevent erroneous reading. For overcoming the problem, the present invention provides a method of controlling the temperature profile of an amorphous silicon in the laser annealing for crystallizing and activating the amorphous silicon thereby controlling the crystal grain boundaries. According to the invention, variation in the electric property of the diode can be decreased and the yield of the phase-change memory can be improved.

Abstract: A semiconductor device is provided with an SRAM cell unit. The SRAM cell unit is provided with a data storing section composed of a pair of drive transistors and a pair of load transistors; a data write section composed of a pair of access transistors; and a data read section composed of an access transistor and a drive transistor. Each of the transistors is provided with a semiconductor layer protruding from a base plane; a gate electrode extending on the both facing side planes over the semiconductor layer from above; a gate insulating film between a gate electrode and a semiconductor layer; and a source/drain region. Each semiconductor layer is arranged to have its longitudinal direction along a first direction. In the adjacent SRAM cell units in the first direction, all the corresponding transistors have the semiconductor layer of one transistor on a center line which is along the first direction of the semiconductor layer of the other transistor.