Abstract: A semiconductor device memory cell (100) can include a built-in capacitor for reducing a soft-error rate (SER). A memory cell (100) can include a first inverter (102) and second inverter (104) arranged in a cross-coupled configuration. A capacitor (110) can be coupled between a first storage node (106) and second storage node (108). A capacitor (110) can be a “built-in” capacitor formed with interconnect wirings utilized to connect memory cell circuit components.

Abstract: A memory capable of reducing the memory cell size is provided. In this memory, a first gate electrode of a first selection transistor and a second gate electrode of a second selection transistor are provided integrally with a word line, and arranged to obliquely extend with respect to the longitudinal direction of a first impurity region on a region formed with memory cells and to intersect with the first impurity region on regions formed with the first selection transistor and the second selection transistor in plan view.

Abstract: An apparatus and method for a multilayer silicon over insulator (SOI) device is provided. In the multilayer SOI device, the crystal orientation of at least one active region of a device is different than the active region of at least another device. Where the multilayer SOI device has a first layer including a PMOS device with a silicon active region having a crystal orientation of [100], the second layer may be an NMOS device with a active region having a silicon layer having a crystal orientation of [110]. The second layer is bonded to the first layer. The method and apparatus can be extended to more than two layers thus forming a multilayer SOI device having a different crystal orientation at each layer. The multiple layer SOI device may form circuits of reduced surface area.

Abstract: A static memory device includes at least one memory cell with two cross-coupled CMOS inverters to be connected to first and second voltages. The substrate of the NMOS transistor of a first CMOS inverter is electrically insulated from the substrate of the NMOS transistor of the second CMOS inverter. The two substrates can be biased with the first voltage. A clear flash controller flash clears the cells for temporarily bring the bias of the substrate of the NMOS transistor of the first CMOS inverter to the second voltage.

Abstract: A dual structure is introduced to the transistor in a flip-flop or a data input step controlled by a clock of a semiconductor logic circuit. The dual structure is formed by connecting a transistor with a MOS transistor having a channel of the same conductivity type in series with respect to the line of a source or drain and connecting their gates to each other, or by connecting an inverter with p-MOS transistors, one for VDD side and one for VSS side of the output step. The dual structure prevents single event phenomenon in a semiconductor logic circuit, such as inverter, SRAM and data latch circuit.

Abstract: A static random access memory (SRAM) cell structure at least comprising a substrate, a transistor, an upper electrode and a capacitor dielectric layer. A device isolation structure is set up in the substrate to define an active region. The active region has an opening. The transistor is set up over the active region of the substrate. The source region of the transistor is next to the opening. The upper electrode is set up over the opening such that the opening is completely filled. The capacitor dielectric layer is set up between the upper electrode and the substrate.

Abstract: A D-Cache SRAM cell having a modified design in schematic and layout that exhibits increased symmetry from the circuit schematic and the physical cell layout perspectives. That is, the SRAM cell includes two read ports and minimizes asymmetry by provisioning one read port on a true side and one on the complement side. Asymmetry is additionally minimized in layout as cross coupling on both the true and complement sides rises up one level by providing from the local interconnect level a via connection to a M1 or metallization level. Moreover, the distance between the local interconnect (MC) and the gate conductor structure (PC) has been enlarged and equalized for each of the pFETs in the cross-latched SRAM cell. As a result, the SRAM cell has been rendered insensitive to overlay (local interconnect processing too close) by maximizing this MC-PC distance.

Abstract: A semiconductor device with improved soft error rate immunity and latch-up immunity and a method of forming the same. The device includes first wells of first conductivity type and second well of second conductivity type formed in the semiconductor substrate of first conductivity type. First conductivity type MOSFETs including source/drain of first conductivity type are formed in the second well, and second conductivity type MOSFETs including source/drain of second conductivity type in the first well. A third well of second conductivity type is formed at a region under the first wells and the drain of the second conductivity type MOSFETs. The first well is connected to the semiconductor substrate between the first well and the third well.

Abstract: The basic cell constituted by a semiconductor integrated circuit comprises two PMOS transistors and two NMOS transistors. By setting the gate widths of the gates of these transistors to prescribed lengths, the efficiency of use of elements within the basic cell is improved, and fine adjustment of the threshold voltage Vth and delay time Tpd becomes possible.

Abstract: A multiport memory cell (200, 300, 600) includes a first word line (WL1) coupled to a gate electrode of a first transistor (201, 301, 601). A second word line (WL2) is coupled to a gate electrode of a second transistor (202, 302, 602). Importantly, the memory cell (200, 300, 600) includes a conductive path (215, 315) between an electrically floating body (426) of the first transistor (201) and an electrically floating body (426) of the second transistor (202). The first word line (WL1) may overlie a first portion of a common body (426) and the second word line (WL2) may overlie a second portion of the common body (426). The common body (426) may be positioned vertically between a buried oxide layer (427) and a gate dielectric layer (430) and laterally between first and second source/drain regions (401, 407) formed in a semiconductor layer (425).

Abstract: Low-power, all-p-channel enhancement-type metal-oxide semiconductor field-effect transistor (PMOSFET) SRAM cells are disclosed. A PMOSFET SRAM cell is disclosed. The SRAM cell can include a latch having first and second PMOSFETs for storing data. Further, a gate of the first PMOSFET is connected to a drain of the second PMOSFET at a first memory node. A gate of the second PMOSFET is connected to a drain of the first PMOSFET at a second memory node. The SRAM cell can also include third and fourth PMOSFETs forming a pull-down circuit. A source of the third PMOSFET is connected to the first memory node. Further, a source of the fourth PMOSFET is connected to the second memory node. The SRAM cell can include access circuitry for accessing data at the first and second memory nodes for read or write operations.

Abstract: A 6T SRAM cell includes a first inverter having a first pull-up transistor and a first pull-down transistor serially coupled between a supply source and a complementary supply source, and a second inverter cross-coupled with the first inverter having a second pull-up transistor and a second pull-down transistor serially coupled between the supply source and the complementary supply source. The cell further includes a first pass-gate and second pass-gate transistors coupled to the first and second inverters, respectively. The first pass-gate transistor and the first pull-up transistor are respectively constructed on a first P-type well and a first N-type well adjacent to one another, which are overlaid by a first doped region and a second doped region of substantially the same width in alignment with one another, respectively.

Abstract: A contact connected to a word line is formed on a gate electrode of an access transistor of an SRAM cell. The contact passes through an element isolation insulating film to reach an SOI layer. A body region of a driver transistor and that of the access transistor are electrically connected with each other through the SOI layer located under the element isolation insulating film. Therefore, the access transistor is in a DTMOS structure having the gate electrode connected with the body region through the contact, which in turn is also electrically connected to the body region of the driver transistor. Thus, operations can be stabilized while suppressing increase of an area for forming the SRAM cell.

Abstract: A memory structure that reduces soft-errors for us in CMOS devices is provided. The memory cell layout utilizes transistors oriented such that the source-to-drain axis is parallel a shorted side of the memory cell. The dimensions of the memory cell are such that it has a longer side and a shorter side, wherein the longer side is preferably about twice as long as the shorter side. Such an arrangement uses a shorter well path to reduce the resistance between transistors and the well strap. The shorter well strap reduces the voltage during operation and soft errors.

Abstract: Disclosed is an improved power grid design for split-word line style memory cell. An array of memory cells comprises a first metal layer for local interconnections; a second metal layer for a bit line, a complementary bit line, and a first voltage line located between the bit line and the complementary bit line; a third metal layer for a first plurality of second voltage lines, and a word line located between the first plurality of second voltage lines, each running substantially in a first direction; and a fourth metal layer for a second plurality of second voltage lines, each running in a second direction orthogonal to the first direction.

Abstract: A method for connecting circuit elements within an integrated circuit for reducing single-event upsets is disclosed. The integrated circuit includes a first and second circuit elements that are substantially identical to each other. In order to reduce the single-event upsets to the first and second circuit elements, each of the first and second circuit elements is divided into a first sub-element and a second sub-element. The first sub-element of the first circuit element is connected to the second sub-element of the second circuit element. The second sub-element of the first circuit element is connected to the first sub-element of the second circuit element. As a result, the nodal spacings between the sub-elements within the first and second circuit elements are effectively increased without demanding additional real estate.

Abstract: A semiconductor integrated circuit device may include a semiconductor substrate, a static memory cell on the semiconductor substrate, a tensile stress film on the pull-down transistors, and a compressive stress film on the pass transistors. The static memory cell may include multiple pull-up transistors and pull-down transistors, which form a latch, and multiple pass transistors may be used to access the latch.

Abstract: A semiconductor memory device includes first and second source/drain regions, and first and second semiconductor regions. The first source/drain region of a first conductive type is formed in a first well region of a second conductive type for a pair of first MIS-type transistors of the first conductive type. The second source/drain region of the second conductive type is formed in a second well region of the first conductive type for a pair of second MIS-type transistors of the second conductive type. The first semiconductor region of the second conductive type is formed in the first source/drain region. The second semiconductor region of the first conductive type is formed in the second source/drain region.

Abstract: A layout structure of a static random access memory (SRAM) cell array includes at least one SRAM cell area, oxide defined (OD) area and strapping cell area. The SRAM cell area has a longitudinal side being at least twice longer than a transverse side thereof. The oxide defined (OD) area is formed on an insulating layer, extending across at least two neighboring SRAM cell areas for construction of a passing gate transistor and a pull-down transistor used in an SRAM cell. The strapping cell area is interposed between the SRAM cell areas, in which a strapping cell is constructed for connecting the OD area to a fixed potential, thereby preventing bodies of the passing gate transistor and the pull-down transistor constructed on the OD area from floating.

Abstract: A semiconductor memory device includes a vertical MISFET having a source region, a channel forming region, a drain region, and a gate electrode formed on a sidewall of the channel forming region via a gate insulating film. In manufacturing the semiconductor memory device, the vertical MISFET in which leakage current (off current) is less can be realized by: counter-doping boron of a conductivity type opposite to that of phosphorus diffused into a poly-crystalline silicon film (10) constituting the channel forming region from an n type poly-crystalline silicon film (7) constituting the source region of the vertical MISFET, and the above-mentioned poly-crystalline silicon film (10); and reducing an effective impurity concentration in the poly-crystalline silicon film (10).

Abstract: An integrated circuit that has logic and a static random access memory (SRAM) array has improved performance by treating the interlayer dielectric (ILD) differently for the SRAM array than for the logic. The N channel logic and SRAM transistors have ILDs with non-compressive stress, the P channel logic transistor ILD has compressive stress, and the P channel SRAM transistor at least has less compressive stress than the P channel logic transistor, i.e., the P channel SRAM transistors may be compressive but less so than the P channel logic transistors, may be relaxed, or may be tensile. It is beneficial for the integrated circuit for the P channel SRAM transistors to have a lower mobility than the P channel logic transistors. The P channel SRAM transistors having lower mobility results in better write performance; either better write time or write margin at lower power supply voltage.

Abstract: A static random access memory (SRAM) device including a substrate and an SRAM unit cell. The substrate includes an n-doped region interposing first and second p-doped regions. The SRAM unit cell includes: (1) a first pass-gate transistor and a first pull-down transistor located at least partially over the first p-doped region; (2) first and second pull-up transistors located at least partially over the n-doped region; and (3) a second pass-gate transistor, a second pull-down transistor, and first and second read port transistors, all located at least partially over the second p-doped region. A boundary of the SRAM unit cell comprises first and second primary dimensions having an aspect ratio of at least about 3.2.

Abstract: For one disclosed embodiment, an apparatus comprises a first p-type device coupled between a cell voltage node and a storage node, an n-type device coupled between the storage node and a reference voltage node, and a second p-type device to couple the storage node to a bit line in response to a signal on a select line. At least one side of diffusion regions in a substrate to form both the first p-type device and the second p-type device are substantially aligned. Other embodiments are also disclosed.

Abstract: To reduce a current loss through a channel and improve electron mobility, a first semiconductor layer and a second semiconductor layer (sequentially formed on a semiconductor substrate) have different lattice properties. The first semiconductor layer and the second semiconductor layer may be etched to form a first semiconductor pattern. A third semiconductor layer having a lattice property substantially identical to that of the first semiconductor layer may be formed over the first semiconductor pattern. The third semiconductor layer may then be etched to form a second semiconductor pattern. A gate may be formed on the second semiconductor pattern. The contact surface between the second semiconductor pattern and the gate pattern may consequently increased to reduce a current loss. Further, the lattice properties may be changed to improve electron mobility of the semiconductor layers.

Abstract: According to some embodiments of the invention, a substrate doped with a P type impurity is provided. An N type impurity is doped into the substrate to divide the substrate into a P type impurity region and an N type impurity region. Active patterns having a first pitch are formed in the P type and N type impurity regions. Gate patterns having a second pitch are formed on the active patterns in a direction substantially perpendicular to the active patterns. Other embodiments are described and claimed.

Abstract: A column select circuit in a Static Random Access Memory (SRAM) having a three-dimensional layout can include a lower CMOS layer in a substrate and an upper NMOS layer above the lower layer. An intermediate PMOS layer is located between the upper NMOS layer and the lower CMOS layer.

Abstract: Semiconductor memory having memory cells, each including first and second conductively-doped contact regions and a channel region arranged between the latter, formed in a web-like rib made of semiconductor material and arranged one behind the other in this sequence in the longitudinal direction of the rib. The rib has an essentially rectangular shape with an upper side of the rib and rib side faces lying opposite. A memory layer is configured for programming the memory cell, arranged on the upper side of the rib spaced apart by a first insulator layer, and projects in the normal direction of the one rib side face over one of the rib side faces so that the one rib side face and the upper side of the rib form an edge for injecting charge carriers from the channel region into the memory layer.

Abstract: Semiconductor devices having a dual polysilicon electrode and a method of manufacturing are provided. The semiconductor devices include a first polysilicon layer deposited on a second polysilicon layer. Each polysilicon layer may be doped individually. The method also allows for some semiconductor devices on a wafer to have a single polysilicon wafer and other devices to have a dual polysilicon layer. In one embodiment, the semiconductor devices are utilized to form a memory device wherein the storage capacitors and transistors located in the cell region are formed with a dual polysilicon layer and devices in the periphery region are formed with a single polysilicon layer.

Abstract: In a complete CMOS SRAM having a memory cell composed of six MISFETs formed over a substrate, a capacitor element having a stack structure is formed of a lower electrode covering the memory cell, an upper electrode, and a capacitor insulating film (dielectric film) interposed between the lower electrode and the upper electrode. One electrode (the lower electrode) of the capacitor element is connected to one storage node of a flip-flop circuit, and the other electrode (the upper electrode) is connected to the other storage node. As a result, the storage node capacitance of the memory cell of the SRAM is increased to improve the soft error resistance.

Abstract: A channel 16 of a FinFET 10 has a channel core 24 and a channel envelope 32, each made from a semiconductor material defining a different lattice structure to exploit strained silicon properties. A gate is coupled to the channel envelope through a gate dielectric. Exemplary materials are Si and SixGe1-x, wherein 78<x<92. The channel core 24 has a top surface 26 of width wc and an upstanding surface 28, 30 of height hc, preferably oriented 90° to one another. The channel envelope 32 is in contact with the top 26 and upstanding surfaces 28, 30 so that the area of interface is increased as compared to contact only along the top surface 26, improving electrical conductivity and gate 18 control over the channel 16. The height hc can be tailored to enable a smaller scale FET 10 within a stabilized SRAM. Various methods of making the channel 16 are disclosed, including a mask and etch method, a handle wafer/carrier wafer method, and a shallow trench method.

Abstract: A semiconductor memory includes first to sixth ridges, an insulating layers on the first to sixth ridges, a first gate line above the first to fourth ridges, and a second gate line above the third to sixth ridges, wherein the first and sixth ridges, the insulating layers, and the first and second gate lines implement first and second capacitors, the second and third ridges and the first gate line implement first driver and load transistors, and the fourth and fifth ridges and the second gate lines implement second load and driver transistors.

Abstract: Bitcell layouts for use in electronic devices and systems are described. One embodiment relates to a memory including at least one bitcell, the bitcell including a storage cell region and a read channel region. The storage cell region is substantially L-shaped and includes six transistors. The read channel region is shaped to fit together with the substantially L-shaped storage cell region to form a substantially rectangular bitcell shape. Other embodiments are described and claimed.

Abstract: Unit cells of a static random access memory (SRAM) are provided including an integrated circuit substrate and first and second active regions. The first active region is provided on the integrated circuit substrate and has a first portion and a second portion. The second portion is shorter than the first portion. The first portion has a first end and a second end and the second portion extends out from the first end of the first portion. The second active region is provided on the integrated circuit substrate. The second active region has a third portion and a fourth portion. The fourth portion is shorter than the third portion. The third portion is remote from the first portion of the first active region and has a first end and a second end. The fourth portion extends out from the second end of the third portion towards the first portion of the first active region and is remote from the second portion of the first active region. Methods of forming SRAM cells are also described.

Abstract: A memory using an SRAM memory cell intended for low-voltage operation is designed to decrease the threshold value of MOS transistors constituting the memory cell without substantial decrease in the static noise margin, which is the operational margin of the memory cell. To this end, a voltage Vdd? higher than a power supply voltage Vdd of a power supply line for peripheral circuits is supplied from a power supply line for memory cells as a power supply voltage for memory cells. Since the conductance of driver MOS transistors is increased, the threshold voltage of the MOS transistors within the memory cells can be reduced without reducing the static noise margin. Further the ratio of width between the driver MOS transistor and a transfer MOS transistor can be set to 1, thereby allowing a reduction in the memory cell area.

Abstract: A method of making a vertical diode is provided, the vertical diode having associated therewith a diode opening extending through an insulation layer and contacting an active region on a silicon wafer. A titanium silicide layer covers the interior surface of the diode opening and contacts the active region. The diode opening is initially filled with an amorphous silicon plug that is doped during deposition and subsequently recrystallized to form large grain polysilicon. The silicon plug has a top portion that is heavily doped with a first type dopant and a bottom portion that is lightly doped with a second type dopant. The top portion is bounded by the bottom portion so as not to contact the titanium silicide layer. For one embodiment of the vertical diode, a programmable resistor contacts the top portion of the silicon plug and a metal line contacts the programmable resistor.

Abstract: A method and system for forming a semiconductor device having superior ESD protection characteristics. A resistive material layer is disposed within a contact hole on at least one of the contact stud upper and lower surface. In preferred embodiments, the integral resistor has a resistance value of between about one Ohm and about ten Ohms, or between 10 and 100 Ohms. Embodiments of the resistive layer include sputtered silicon material, a tunnel oxide, a tunnel nitride, a silicon-implanted oxide, a silicon-implanted nitride, or an amorphous polysilicon. Embodiments of the invention include SRAMs, bipolar transistors, SOI lateral diodes, MOSFETs and SiGe Transistors.

Abstract: A semiconductor device includes a cross diffusion barrier layer sandwiched between a gate layer and an electrode layer. The gate layer has a first gate portion of doped polysilicon of first conductivity type adjacent to a second gate portion doped polysilicon of second conductivity type. The cross diffusion barrier layer includes a combination of silicon and nitrogen. The cross diffusion barrier layer adequately prevents cross diffusion between the first and second gate portions while causing no substantial increase in the resistance of the gate layer.

Abstract: A static random access memory (SRAM) cell structure at least comprising a substrate, a transistor, an upper electrode and a capacitor dielectric layer. A device isolation structure is set up in the substrate to define an active region. The active region has an opening. The transistor is set up over the active region of the substrate. The source region of the transistor is next to the opening. The upper electrode is set up over the opening such that the opening is completely filled. The capacitor dielectric layer is set up between the upper electrode and the substrate.

Abstract: SRAM cells having landing pads in contact with upper and lower cell gate patterns, and methods of forming the same are provided. The SRAM cells and the methods remove the influence resulting from structural characteristics of the SRAM cells having vertically stacked upper and lower gate patterns, for stably connecting the patterns on the overall surface of the semiconductor substrate. An isolation layer isolating at least one lower active region is formed in a semiconductor substrate of the cell array region. The lower active region has two lower cell gate patterns. A body pattern is disposed in parallel with the semiconductor substrate. The body pattern is formed to confine an upper active region, which has upper cell gate patterns on the lower cell gate patterns. A landing pad is disposed between the lower cell gate patterns. A node pattern is formed to simultaneously contact the upper cell gate pattern and the lower cell gate pattern.

Abstract: A semiconductor integrated circuit device is configured by eight transistors including the six transistors configuring the data holding section and the two NMOS transistors configuring the reading stage. The threshold voltage of the NMOS transistors configuring the reading stage is set low and the threshold voltage of the six transistors configuring the data holding section is set higher than the threshold voltage of the NMOS transistors configuring the reading stage. The cell current flowing from the bit line to the ground terminal can be set large and the large static noise margin (SNM) can be attained.

Abstract: An integrated circuit is disclosed having one or more devices having substantially similar physical gate electric thicknesses but different electrical gate electric thicknesses for accommodating various operation needs. One or more devices are manufactured with a same mask set using multiple doping processes to generate substantially similar physical gate dielectric thicknesses, but with different electrical gate dielectric thicknesses. The device undergoing multiple doping processes have different dopant concentrations, thereby providing different electrical characteristics such as the threshold voltages.

Abstract: Ternary CAM cells are provided. The ternary CAM cell includes a pair of half cells. Each of the half cells includes an isolation layer formed at a predetermined region of a semiconductor substrate to define a match cell active region. A search gate electrode and a node gate electrode are placed to cross over the match cell active region. A match line is electrically connected to the match cell active region, which is adjacent to the node gate electrode and is located opposite the search gate electrode. An SRAM cell is provided at the semiconductor substrate adjacent to the match cell active region. The node gate electrode is electrically connected to one of a pair of storage nodes of the SRAM cell.

Abstract: In a semiconductor memory device, third and fourth transistors are configured as a vertical structure. The third transistor is laminated over a first transistor, and the fourth transistor is laminated over a second transistor, whereby a reduction in cell area is achieved. A voltage, which is set on the condition that the difference between a source potential applied to each of the first and second transistors and the potential of a select level of a word line becomes greater than or equal to a threshold voltage of each of the third and fourth transistors, is supplied to a source electrode of each of the first and second transistors, to thereby perform “0” write compensation.

Abstract: The present invention describes methods, apparatus, and systems related to polysilicon gate contact openings over active regions formed by a separate mask to provide enough control of dielectric removal to produce a contact opening at least down to the gate layer but not down to the junction layers. Embodiments include, self-aligned polysilicon contacts done by timed contact etch, by a two layer dielectric, by adding a dielectric etch stop layer, and by partially planarizing a dielectric or etch stop layer over the gate layer. Thus, even if mis-aligned, the gate contact openings will be deep enough to reach active region gates, but not deep enough to reach junctions. As a result, by using a separate mask and by selecting a period of time for etching to active gates, gate contact openings can be formed during manufacture of ICs, semiconductors, MOS memory cells, SRAM, flash memory, and various other memory cells.

Abstract: A compact semiconductor structure having back gate(s) for controlling threshold voltages and associated method of formation is disclosed. Fabrication of the semiconductor structure starts with a semiconductor region formed directly on an underlying electrically isolating layer. Then, a mandrel and a spacer are formed on the semiconductor region. Next, a back gate region is formed separated from the semiconductor region by a back gate isolating layer and covered by an inter-gate isolating layer. Next, a portion of the semiconductor region beneath the mandrel is removed so as to form an active region adjacent to the removed portion of the semiconductor region. Finally, a main gate region is formed in place of the removed portion of the semiconductor region and on the inter-gate isolating layer. The main gate region is separated from the active region by a main gate isolating layer and separated from the back gate region by the inter-gate isolating layer.

Abstract: In a full CMOS SRAM having a lateral type cell (memory cell having three partitioned wells arranged side by side in a word line extending direction and longer in the word line direction than in the bit line direction) including first and second driver MOS transistors, first and second load MOS transistors and first and second access MOS transistors, two capacitors are arranged spaced apart from each other on embedded interconnections to be storage nodes, with lower and upper cell plates cross-coupled to each other.

Abstract: A configuration is provided to reduce variations in the width of the gate of a read-out transistor without increasing the surface area of a memory cell. To do this, a recess is provided in an inner corner of a gate electrode that is bent into an L-shape. The recess is located so as to face a rectangular portion of an active region of the memory cell.

Abstract: A semiconductor device according to the present invention includes: a semiconductor substrate including an active region and an isolation region; a gate electrode formed on the active region with an oxide film interposed therebetween; and a set of impurity regions formed on both sides of the gate electrode. A surface of the active region is entirely rounded so as to be inclined downward toward the isolation region. This rounded shape can be formed by forming an isolation oxide film such that a bird's beak portion is connected on the active region.

Abstract: A semiconductor device has a first and a second semiconductor layer provided on an insulating film on a support substrate. A first memory cell transistor, which constitutes a part of a memory cell in an SRAM, has a first gate electrode of a first conductivity type and first source/drain diffusion layers of a second conductivity type opposite to the first conductivity type. The following expression is fulfilled the thickness of the first conductivity type?one-third of a length of the first gate electrode in its channel length. A first peripheral transistor, which constitutes a part of a peripheral circuit, has a third gate electrode and a third source/drain diffusion layers. The following expression is satisfied the thickness of the second semiconductor layer>one-third of a length of the third gate electrode in its channel length direction.

Abstract: A static random access memory has first and second complementary field-effect transistors. The first complementary field-effect transistor includes a semiconductor substrate, a first field-effect transistor of electron conduction type which has a first drain region constituting a Schottky junction and a gate electrode, and a first field-effect transistor of positive hole conduction type which shares the first drain region and has a shared gate electrode. The second complementary field-effect transistor includes a second field-effect transistor of electron conduction type which has a second drain region and a gate electrode, a second field-effect transistor of positive hole conduction type which shares the second drain region and has a shared gate electrode.