Abstract: A method of forming an integrated circuit includes forming first structures in a first portion of the integrated circuit and forming second structures, which are arranged more densely than the first structures, in a second portion. The first and second structures are defined by lithography processes using photomasks. At least one of the photomasks includes both openings in a first region for supporting the definition of the first structures and openings in a second region for supporting the definition of the second structures.

Abstract: The present invention has for its purpose to provide a technique capable of reducing planar dimension of the semiconductor device. An input/output circuit is formed over the semiconductor substrate, a grounding wiring and a power supply wiring pass over the input/output circuit, and a conductive layer for a bonding pad is formed thereover. The input/output circuit is formed of MISFET elements in the nMISFET forming region and the pMISFET forming region, resistance elements in the resistance element forming regions and diode elements in the diode element forming regions functioning as protective elements. A wiring connected to the protective elements and positioned under the grounding wiring and the power supply wiring is pulled out in a pulling-out region between the nMISFET forming region and the pMISFET forming region and between the grounding wiring and the power supply wiring to be connected to the conductive layer.

Abstract: A power and ground shield mesh to remove both capacitive and inductive signal coupling effects of routing in integrated circuit device. An embodiment describes the routing of a shield mesh of both power and ground lines to remove noise created by capacitive and inductive coupling. Relatively long signal lines are routed in between fully connected power and ground shield mesh which may be generated by a router during the signal routing phase or during power mesh routing phase. Leaving only the odd tracks or the even tracks for signal routing, power mesh (VDD) and ground mesh (VSS) are routed and fully interconnected leaving shorter segments and thereby reducing the RC effect of the circuit device. Another embodiment presents a technique where the signals are shielded using the power and ground mesh for a gridless routing. Another embodiment presents a multi-layer grid routing technique where signals are routed on even grid and the power and ground lines are routed on odd grid.

Abstract: CMOS inverters are included in a standard cell. Power supply lines are electrically connected to CMOS inverters, and include lower layer interconnects and upper layer interconnect. Lower layer interconnects extend along a boundary of standard cells adjacent to each other and on the boundary. Upper layer interconnects are positioned more inside in standard cell than lower layer interconnects, as viewed from a plane. CMOS inverters are electrically connected through upper layer interconnects to lower layer interconnects. Thus, a semiconductor device is obtained that can achieve both higher speeds and higher integration.

Abstract: A filler cell for use in fabricating an integrated circuit. The filler cell couples a power supply rail of an adjacent logic cell to a power supply rail of another adjacent logic cell. The filler cell also has a diode to bleed charge accumulated on the power rails of the adjacent logic cells to the substrate. The diode is reverse biased during normal integrated circuit operation. A method for fabricating an integrated circuit with a power grid. At least one filler cell is placed on the integrated circuit to bleed away charge accumulated on the power grid during the fabrication of the integrated circuit. The filler cell is connected to a supply rail of an adjacent logic cell.

Abstract: Disclosed is a microelectromechanical system (MEMS) device and method of manufacturing the same. In one aspect, MEMS such as an interferometric modulator include one or more elongated interior posts and support rails supporting a deformable reflective layer, where the elongated interior posts are entirely within an interferometric cavity and aligned parallel with the support rails. In another aspect, the interferometric modulator includes one or more elongated etch release holes formed in the deformable reflective layer and aligned parallel with channels formed in the deformable reflective layer defining parallel strips of the deformable reflective layer.

Abstract: A unit cell for an integrated circuit includes a first conductive type active region and a second conductive type active region which extend in a first direction. Each of the active regions has first and second ends. The first end of the second conductive type active region opposes the second end of the first conductive type active region. A poly-silicon pattern extends in the first direction across the first conductive type active region and second conductive type active region. A first contact region is adjacent the first end of the first conductive type active region in the first direction. A second contact region is adjacent the second end of the second conductive type active region in the first direction.

Abstract: A semiconductor memory device includes a semiconductor substrate having a dummy cell region adjacent to a memory cell region, a plurality of memory cell transistors, a selective gate transistor, a peripheral circuit transistor, a selective gate line, a contact plug, a dummy contact plug formed in an element forming region of the memory cell region adjacent to the selective gate line, and a spacer insulating film formed on a sidewall of the peripheral circuit transistor. The sidewall of the selective gate electrode is formed with no spacer insulating film, and the selective gate line has a sidewall facing an region of the dummy cell region in which the dummy contact plug is formed, except for the sidewall of the selective gate electrode. The sidewall of the selective gate line is formed with a spacer insulating film.

Abstract: Provided is a semiconductor device that is capable of suppressing occurrence of a crystal defect in an elongated circuit region formed in an SOI substrate. Low-voltage transistor regions are separated, by multiple inner isolation layers, into multiple sub-regions. For this reason, the length of the longitudinal direction of the sub-regions is reduced, even though the low-voltage transistor regions are extremely elongated, for example. This configuration can suppress occurrence of a crystal defect in the low-voltage transistor regions in the longitudinal direction thereof, although such defect may occur due to the difference in thermal expansion or thermal contraction between a semiconductor layer in the low-voltage transistor regions, and the element isolation layers.

Abstract: The present invention provides an integrated circuit suitable for various packaging modes. This integrated circuit includes: a core circuit, a plurality of pads, and a selection circuit. The selection circuit is coupled between the core circuit and the pads for determining the connection state between the core circuit and the pads based on a control signal. When the control signal provides a first value, the core circuit and the pads will be in a first connection state, and the integrated circuit will be applied with a single-die package. However, when the control signal provides a second value, the core circuit and the pads will be in the second connection state, and the integrated circuit will be applied with a multi-die package.

Abstract: An imager element, device and imaging system image sensor pixel. The image sensor pixel includes a collection region, a floating diffusion region, and a transfer transistor having a recessed gate. The recessed gate is configured to couple the collection region to the floating diffusion region so that collected charge is transferred during activation. The recessed gate has an effective gate length greater than the physical gate length.

Abstract: A semiconductor device, formed on a semiconductor substrate, including a first memory array formed in a first region and including first word lines, first bit lines across the first word lines, and memory cells at intersections of the first word lines and the first bit lines, a second memory array which is formed in a second region and including second word lines, second bit lines across the second word lines, and memory cells at intersections of the second word lines and the second bit lines, and address pads located in a third region, in which the first region, the third region and the second region are arranged in that order in the first direction, the address input pads being arranged between a center axis of the first direction of the substrate and the first region, and no address input pads are arranged between the center axis and the second region.

Abstract: A semiconductor memory device and a manufacturing method thereof are provided which enable cell-contact plugs to be formed at high yields and the yields of semiconductor memory devices to be improved in the manufacturing process. The semiconductor memory device includes: a semiconductor substrate; MOS transistors which are formed on a surface of the semiconductor substrate; a cell-contact plug which is made of poly-silicon film, is located between gates of the MOS transistors, and is connected to a source or a drain of one of the MOS transistors; a pad metal layer which is formed on the cell-contact plug; an interlayer dielectric film which is formed on the pad metal layer; a storage capacitor which is formed on the interlayer dielectric film; and a contact plug which is formed inside an opening which penetrates the interlayer dielectric film, and connects the storage capacitor with the pad metal layer.

Abstract: The Vertical System Integration (VSI) invention herein is a method for integration of disparate electronic, optical and MEMS technologies into a single integrated circuit die or component and wherein the individual device layers used in the VSI fabrication processes are preferably previously fabricated components intended for generic multiple application use and not necessarily limited in its use to a specific application. The VSI method of integration lowers the cost difference between lower volume custom electronic products and high volume generic use electronic products by eliminating or reducing circuit design, layout, tooling and fabrication costs.

Abstract: A rewritable nonvolatile memory cell is disclosed comprising a steering element in series with a carbon nanotube fabric. The steering element is preferably a diode, but may also be a transistor. The carbon nanotube fabric reversibly changes resistivity when subjected to an appropriate electrical pulse. The different resistivity states of the carbon nanotube fabric can be sensed, and can correspond to distinct data states of the memory cell. A first memory level of such memory cells can be monolithically formed above a substrate, a second memory level monolithically formed above the first, and so on, forming a highly dense monolithic three dimensional memory array of stacked memory levels.

Abstract: An integrated circuit includes N plane-like metal layers. A first plane-like metal layer includes M contact portions that communicate with the N plane-like metal layers, respectively. The first plane-like metal layer and the N plane-like metal layers are located separate planes. First and second drain regions have a symmetric shape across at least one of horizontal and vertical centerlines. First and second gate regions have a first shape that surrounds the first and second drain regions, respectively. First and second source regions are arranged adjacent to and on one side of the first gate region, the second gate region and the connecting region. The first source region, the second source region, the first drain region and the second drain region communicate with at least two of the N plane-like metal layers.

Abstract: Methods and systems for optimal decoupling capacitance in a dual-voltage power-island architecture. In low-voltage areas of the chip, accumulation capacitors of two different types are used for decoupling, depending on whether the capacitor is located in an area which is always-on or an area which is conditionally powered.

Abstract: A technique of manufacturing a semiconductor device in which etching in formation of a contact hole can be easily controlled is proposed. A semiconductor device includes at least a semiconductor layer formed over an insulating surface; a first insulating layer formed over the semiconductor layer; a gate electrode formed over the first insulating layer; a second insulating layer formed over the gate electrode; and a conductive layer formed over the second insulating layer connected to the semiconductor layer via an opening which is formed at least in the semiconductor layer and the second insulating layer and partially exposes the insulating surface. The conductive layer is electrically connected to the semiconductor layer at the side surface of the opening which is formed in the semiconductor layer.

Abstract: A method for easily manufacturing a semiconductor device in which variation in thickness or disconnection of a source electrode or a drain electrode is prevented is proposed A semiconductor device includes a semiconductor layer formed over an insulating substrate; a first insulating layer formed over the semiconductor layer; a gate electrode formed over the first insulating layer; a second insulating layer formed over the gate electrode; an opening which reaches the semiconductor layer and is formed at least in the first insulating layer and the second insulating layer; and a step portion formed at a side surface of the second insulating layer in the opening.

Abstract: A semiconductor integrated circuit is provided which entails no increase in the correction time of OPC and in which non-uniformity in the gate lengths due to the optical proximity effects is surely suppressed. A plurality of standard cells (C1, C2, C3, . . . ), each including gates G extended in the vertical direction, are aligned in the transverse direction to form a standard cell row. A plurality of the standard cell rows are located side by side in the vertical direction to form a standard cell group. Each of the standard cell rows has a terminal standard cell Ce at least one end of the standard cell row. The terminal standard cell Ce includes two or more supplementary gates, each of which is any of a dummy gate and a gate of an inactive transistor.

Abstract: Repeaters are arranged at arbitrary positions to substantially improve transmission speed of a signal. In the semiconductor integrated circuit device 1, repeater regions 10 where repeaters are provided as relay points for wiring are provided in the central parts of the core power source regions 2, 3 and 5, on the left side of the core power source regions 4 to 8 and at the upper and lower parts of the semiconductor integrated circuit device 1. A power switch region for repeater 11 is formed so as to surround the core power source regions 2 to 8 and the repeater regions 10. The power source lines of the reference potential connected to the repeater regions 10 are laid out at equally spaced intervals throughout the core power source regions 2 to 8, which enables the repeater regions 10 to be flexibly laid out. This permits the repeaters to be more effectively arranged, which improves the performances of semiconductor integrated circuit device 1.

Abstract: For ensuring the complete turn-off state of an ESD protecting device and preventing leakage current from a chip, an alternative conducting path is formed in the chip for bypassing an external current. The chip further includes an internal circuit and a conducting circuit.

Abstract: The present invention provides structures for antifuses that utilize electromigration for programming. By providing a portion of antifuse link with high resistance without conducting material and then by inducing electromigration of the conducting material into the antifuse link, the resistance of the antifuse structure is changed. By providing a terminal on the antifuse link, the change in the electrical properties of the antifuse link is detected and sensed. Also disclosed are an integrated antifuse with a built-in sensing device and a two dimensional array of integrated antifuses that can share programming transistors and sensing circuitry.

Abstract: A method of forming an isolation layer of a semiconductor device includes the steps of forming a gate insulating layer and a conductive layer on an active area of a semiconductor substrate; forming a spacer layer on side walls of the conductive layer; forming a trench on the semiconductor substrate between the spacer layer-covered side walls; removing the spacer layer to form a step on an upper edge of the trench; and forming a liner insulating layer on the trench. The method makes it possible to solve problems caused by impurities present in material with which the trench is gap-filled or present in etchants used in an etch-back process.

Abstract: A semiconductor device includes: a gate electrode formed above a semiconductor region; a drain region and a source region formed in portions of the semiconductor region located below sides of the gate electrode in a gate length direction, respectively; a plurality of drain contacts formed on the drain region to be spaced apart in a gate width direction of the gate electrode; and a plurality of source contacts formed on the source region to be spaced apart in the gate width direction of the gate electrode. The intervals between the drain contacts are greater than the intervals between the source contacts.

Abstract: In accordance with an embodiment of the invention, a superjunction semiconductor device includes an active region and a termination region surrounding the active region. A central vertical axis of a boundary column of a second conductivity type material defines the boundary between the active region and the termination region. The active and termination regions include columns of first and second conductivity type material alternately arranged along a horizontal direction in a semiconductor region having top and bottom surfaces. At least one of the columns of the first conductivity type material in the termination region has a different width than a width of the columns of the first conductivity type material in the active region.

Abstract: A demultiplexer using transistors for accessing memory cell arrays. The demultiplexer includes (a) a substrate; (b) 2N semiconductor regions which are parallel to one another and run in a first direction; (c) first N gate electrode lines, which (i) run in a second direction which is perpendicular to the first direction, (ii) are electrically insulated from the 2N semiconductor regions, and (iii) are disposed between the first plurality of memory cells and the contact region; (d) a contact region; (e) a first plurality of memory cells. An intersection transistor exists at each of intersections between the first N gate electrode lines and the 2N semiconductor regions. In response to pre-specified voltage potentials being applied to the contact region and the first N gate electrode lines, memory cells of the first plurality of memory cells disposed on only one of the 2N semiconductor regions are selected.

Abstract: An integrated circuit assembly (ICA) comprises: a digital and/or analog integrated circuit (S1) having a core with input and/or output pins and at least one power supply connection pad (PP) and one ground connection pad (GP) connected to a chosen one of the input and/or output pins and respectively connected to power supply and ground connection zones (MZ1) of a printed circuit board (PCB), and a passive integration substrate (S2) set on top of the digital and/or analog integrated circuit (S1) and comprising i) at least first and second input zones respectively connected to the ground (GP) and power supply (PP) connection pads to, be fed with input ground and supply voltages, ii) input and/or output zones connected to chosen core input and/or output pins, and Ëi) a passive integrated circuit (PIC) connected to the first and second input zones and arranged to feed the substrate input and/or output zones with chosen ground and supply voltages defined from the input ground and supply voltages.

Abstract: There is provided a technology which allows sufficient protection of internal circuits from electrostatic discharge even when internal-circuit power source pads and internal-circuit GND pads are formed on an internal circuit region. Internal-circuit power source pads and internal-circuit GND pads are placed in the core region of a semiconductor chip. Between the internal-circuit power source pads and the internal-circuit GND pads, the internal circuits are formed. Between the internal-circuit power source pads and the internal-circuit GND pads, electrostatic protection circuits for protecting the internal circuits from a surge current are further formed. Each of the electrostatic protection circuits is composed of a discharge circuit for causing the surge current to flow therein and a control circuit for controlling the discharge circuit. The present invention is characterized in that the discharge circuits are placed in the core region and the control circuits are placed in an I/O region.

Abstract: In a layout structure of a semiconductor integrated circuit, when transistors are arranged in a constant gate wiring pitch, a common source diffusion region is provided between two adjacent transistors, a CA via is provided on the common source diffusion region, and a source wiring connected to the CA via is provided on the common source diffusion region. An inter-drain wiring connecting the drain regions of the two transistors is formed in a wiring layer higher than the source wiring. Therefore, the wiring path of the source wiring is not limited by the wiring path of the inter-drain wiring, and can be provided, covering the common source diffusion region to a further extent. As a result, the number of high-resistance CA vias or the flexibility of arrangement is increased, leading to a reduction in source resistance, resulting in an increase in operating speed of the semiconductor integrated circuit.

Abstract: A memory system having electromechanical memory cells and decoders is disclosed. A decoder circuit selects at least one of the memory cells of an array of such cells. Each cell in the array is a crossbar junction at least one element of which is a nanotube or a nanotube ribbon. The decoder circuit is constructed of crossbar junctions at least one element of each junction being a nanotube or a nanotube ribbon.

Abstract: When dummy patterns are arranged to planarize LSI layout patterns, a plurality of dummy patterns 1 are arranged in a wiring layer in which signal wiring patterns 2 are formed, so as to be inclined at an angle of generally 45 degrees toward the associated signal wiring patterns 2. These dummy patterns 1 cross signal wiring patterns 3 formed in another vertically adjacent wiring layer to have an inclination angle of generally 45 degrees. A plurality of dummy patterns 13 are located in the wiring layer in which the signal wiring patterns 3 are formed, so as to be inclined at an angle of generally 45 degrees toward the associated signal wiring patterns 3. The dummy patterns 1 formed in one of the adjacent wiring layers cross the dummy patterns 13 formed in the other wiring layer at an angle of generally 90 degrees. This reduces fluctuations in wiring capacitance and equalizes fluctuations in the wiring capacitance to the maximum extent.

Abstract: Under one aspect, a non-volatile nanotube diode device includes first and second terminals; a semiconductor element including a cathode and an anode, and capable of forming a conductive pathway between the cathode and anode in response to electrical stimulus applied to the first conductive terminal; and a nanotube switching element including a nanotube fabric article in electrical communication with the semiconductive element, the nanotube fabric article disposed between and capable of forming a conductive pathway between the semiconductor element and the second terminal, wherein electrical stimuli on the first and second terminals causes a plurality of logic states.

Abstract: Under one aspect, a nonvolatile nanotube diode includes: a substrate; a semiconductor element disposed over the substrate, the semiconductor element having an anode and a cathode and capable of forming an electrically conductive pathway between the anode and the cathode; a nanotube switching element disposed over the semiconductor element, the nanotube switching element including a conductive contact and a nanotube fabric element capable of a plurality of resistance states; and a conductive terminal disposed in spaced relation to the conductive contact, wherein the nanotube fabric element is interposed between and in electrical communication with the conductive contact and the conductive contact is in electrical communication with the cathode, and wherein in response to electrical stimuli applied to the anode and the conductive terminal, the nonvolatile nanotube diode is capable of forming an electrically conductive pathway between the anode and the conductive terminal.

Abstract: In a semiconductor integrated circuit device, a plurality of electrode pads for external connection are arranged in a zigzag pattern. Some electrode pads of the electrode pads of the plurality of I/O cells which are closer to a side of the semiconductor chip, each have an end portion closer to the side of the semiconductor chip, the end portion being set at the same position as that of an end portion of the corresponding I/O cell. A power source-side protective circuit and a ground-side protective circuit against discharge of static electricity are provided with the power source-side protective circuit being closer to the scribe region. A distance between a center position of one of the electrode pads and the ground-side protective circuit of the corresponding I/O cell and a distance between a center position of the other one electrode pad and the ground-side protective circuit of the corresponding I/O cell are both short and are substantially equal between each I/O cell.

Abstract: Buffers are arranged in a concentrated manner in a region distant from pads. The region refers to a region in a main region of a semiconductor integrated circuit, except for a central processing unit, a non-volatile memory and a volatile memory. As the buffer requiring a large area is not provided around the pad, a pitch between the pads or a pitch between the pad and an internal circuit (such as the central processing unit) can be made smaller and hence a chip size can be reduced. Therefore, a semiconductor integrated circuit capable of achieving a reduced chip size can be provided.

Abstract: The present invention provides an integrated circuit for providing ESD protection. The integrated circuit comprises a transistor device having at least one interleaved finger having a substrate region, a source, drain and gate region formed over a channel region disposed between the source and the drain regions. The transistor device further comprises at least one highly doped junction formed adjacent to the source region to measure voltage potential of the substrate region. The integrated circuit further comprises a switching circuit coupled to the at least one highly doped junction such that the voltage potential is transferred to the switching circuit to either draw the full ESD current or trigger to draw the full ESD current.

Abstract: A method of forming an array of floating gate memory cells, and an array formed thereby, wherein each memory cell includes a substrate of semiconductor material having a first conductivity type, source and drain regions formed in the substrate, a block of conductive material disposed over and electrically connected to the source, and a floating gate having a first portion disposed over and insulated from the source region and a second portion disposed over and insulated from the channel region. The floating gate first portion includes a sloped upper surface and a side surface that meet at an acute edge. An electrically conductive control gate is disposed over and insulated from the channel region for controlling a conductivity thereof.

Abstract: An array of vertical transistor cells formed in a substrate for selecting one of a plurality of memory cells by selecting a word line and a bit line is disclosed. In one embodiment, for minimizing the area of a cell and reducing complexity in production a plurality of parallel insulating trenches filled with an insulating material and a plurality of perpendicular gate electrode trenches is formed, the gate electrode trenches filled with a suitable gate electrode material disrupted by the insulating material thus forming separate gate electrodes arranged below the reference plane. The insulating trenches and the gate electrode trenches form distinct active areas of transistors in the substrate, wherein two gate electrodes located at opposing sidewalls of an active area form a double gate electrode of a transistor, and wherein a plurality of gate electrodes is coupled to a word line running perpendicular to the gate electrode trenches and above the reference plane.

Abstract: A phase-change material is proposed for coupling interconnect lines an electrically programmable matrix array. Leakage may be reduced by optionally placing a thin insulating breakdown layer between the phase change material and at least one of the lines. The matrix array may be used in a programmable logic device. The logic portions of the programmable logic device may be tri-stated.

Abstract: In a semiconductor integrated circuit, a cell arrangement area is provided on a semiconductor substrate to allow a plurality of basis cells to be arranged. A basic power supply line is provided in an upper layer than the cell arrangement area to supply a power. A switch cell is configured to control the power supply from the basic power supply line to an inside of the cell arrangement area. An always operating cell is arranged in the cell arrangement area adjacently to the switch cell, and is configured to receive the power from the switch cell even when the switch cell stops the power supply to the cell arrangement area.

Abstract: A method of constructing an integrated circuit involves selecting modular tiles and then generating a functional circuit layout using the tiles. Modular tiles that perform predetermined functions and that have approximately the same length and width dimensions are selected from a library of validated tiles. The tiles have input-output terminals embedded in their upper active layers. A functional circuit layout for the integrated circuit is generated using the tiles. In many implementations, the physical layout of the integrated circuit does not include the step of routing. Then an interconnect layer is added over the functional circuitry of the tiles and connects the input-output terminals to bond pads located at the perimeter of the functional circuit layout. Chip data corresponding to the functional circuit layout is generated, and then mask reticles corresponding to the chip data are generated. The integrated circuit is formed on a wafer based on the mask reticles.

Abstract: The height H of several kinds of basic cell are made the same and several kinds of macro cell which have a length which is an integral multiplication of the height H of this basic cell, are prepared, the basic cell and macro cell are mixed and the circuit of a peripheral circuit is designed. A M0 wire of a first wiring layer which is formed on a semiconductor substrate is used as a wire used within a macro cell. The basic cell and the macro cell are connected by a M1 wire of a second wiring layer which is formed on the first wiring layer and a M2 wire M2 of a third wiring layer. The transistor layout of basic cells and macro cells is designed and verified in advance and stored in a cell library, and auto routing by a standard method may be carried out.

Abstract: A DRAM fabricated on an SOI substrate employing single body devices as memory cells without relying on a field through the insulative layer of the SOI is described. Floating body devices are defined by orthogonally disposed lines with both a front gate and back gate for each body being formed on the insulative layer.

Abstract: In a memory device and a method of forming the same, in one embodiment, the memory device comprises a first word line structure on a substrate, the first word line structure extending in a first direction. A bit line is provided over the first word line structure and spaced apart from the first word line by a first gap, the bit line extending in a second direction transverse to the first direction. A second word line structure is provided over the bit line and spaced apart from the bit line by a second gap, the second word line structure extending in the first direction.

Abstract: By forming MOSFETs on a substrate having pre-existing ridges of semiconductor material (i.e., a “corrugated substrate”), the resolution limitations associated with conventional semiconductor manufacturing processes can be overcome, and high-performance, low-power transistors can be reliably and repeatably produced. Forming a corrugated substrate prior to actual device formation allows the ridges on the corrugated substrate to be created using high precision techniques that are not ordinarily suitable for device production. MOSFETs that subsequently incorporate the high-precision ridges into their channel regions will typically exhibit much more precise and less variable performance than similar MOSFETs formed using optical lithography-based techniques that cannot provide the same degree of patterning accuracy. Additional performance enhancement techniques such as pulse-shaped doping and “wrapped” gates can be used in conjunction with the segmented channel regions to further enhance device performance.

Abstract: A high power, high frequency semiconductor device has a plurality of unit cells connected in parallel. The unit cells each having a controlling electrode and first and second controlled electrodes. A thermal spacer divides at least one of the unit cells into a first active portion and a second active portion, spaced apart from the first potion by the thermal spacer. The controlling electrode and the first and second controlled electrodes of the unit cell cross over the first thermal spacer.