Abstract: Nanostructure field-effect transistors (NSFETs) including isolation layers formed between epitaxial source/drain regions and semiconductor substrates and methods of forming the same are disclosed. In an embodiment, a semiconductor device includes a semiconductor substrate; a gate stack over the semiconductor substrate, the gate stack including a gate electrode and a gate dielectric layer; a first epitaxial source/drain region adjacent the gate stack; and a high-k dielectric layer extending between the semiconductor substrate and the first epitaxial source/drain region, the high-k dielectric layer contacting the first epitaxial source/drain region, the gate dielectric layer and the high-k dielectric layer including the same material.

Abstract: A method includes forming a first semiconductor fin and a second semiconductor fin over a substrate; forming a first gate structure over the substrate and crossing the first semiconductor fin; forming a second gate structure over the substrate and crossing the second semiconductor fin; forming a first gate spacer on a sidewall of the first gate structure; and forming a second gate spacer on a sidewall of the second gate structure, wherein in a top view, an outer sidewall of the first gate spacer farthest from the first gate structure is coterminous with an outer sidewall of the second gate spacer farthest from the second gate structure, and an inner sidewall of the first gate spacer in contact with the first gate structure is misaligned with an inner sidewall of the second gate spacer in contact with the second gate structure.

Abstract: The present disclosure describes a memory structure including a memory cell array. The memory cell array includes memory cells and first n-type wells extending in a first direction. The memory structure also includes a second n-type well formed in a peripheral region of the memory structure. The second n-type well extends in a second direction and is in contact with a first n-type well of the first n-type wells. The memory structure further includes a pick-up region formed in the second n-type well. The pick-up region is electrically coupled to the first n-type well of first n-type wells.

Abstract: Semiconductor devices having improved source/drain features and methods for fabricating such are disclosed herein. An exemplary device includes a semiconductor layer stack disposed over a mesa structure of a substrate. The device further includes a metal gate disposed over the semiconductor layer stack and an inner spacer disposed on the mesa structure of the substrate. The device further includes a first epitaxial source/drain feature and a second epitaxial source/drain feature where the semiconductor layer stack is disposed between the first epitaxial source/drain feature and the second epitaxial source/drain feature. The device further includes a void disposed between the inner spacer and the first epitaxial source/drain feature.

Abstract: SRAM designs based on GAA transistors are disclosed that provide flexibility for increasing channel widths of transistors at scaled IC technology nodes and relax limits on SRAM performance optimization imposed by FinFET-based SRAMs. GAA-based SRAM cells described have active region layouts with active regions shared by pull-down GAA transistors and pass-gate GAA transistors. A width of shared active regions that correspond with the pull-down GAA transistors are enlarged with respect to widths of the shared active regions that correspond with the pass-gate GAA transistors. A ratio of the widths is tuned to obtain ratios of pull-down transistor effective channel width to pass-gate effective channel width greater than 1, increase an on-current of pull-down GAA transistors relative to an on-current of pass-gate GAA transistors, decrease a threshold voltage of pull-down GAA transistors relative to a threshold voltage of pass-gate GAA transistors, and/or increases a ? ratio of an SRAM cell.

Abstract: Semiconductor device and the manufacturing method thereof are disclosed. An exemplary semiconductor device comprises a first semiconductor stack and a second semiconductor stack over a substrate, wherein each of the first and second semiconductor stacks includes semiconductor layers stacked up and separated from each other; a dummy spacer between the first and second semiconductor stacks, wherein the dummy spacer contacts a first sidewall of each semiconductor layer of the first and second semiconductor stacks; and a gate structure wrapping a second sidewall, a top surface, and a bottom surface of each semiconductor layer of the first and second semiconductor stacks.

Abstract: A method of fabricating a device includes providing a fin extending from a substrate in a device type region, where the fin includes a plurality of semiconductor channel layers. In some embodiments, the method further includes forming a gate structure over the fin. Thereafter, in some examples, the method includes removing a portion of the plurality of semiconductor channel layers within a source/drain region adjacent to the gate structure to form a trench in the source/drain region. In some cases, the method further includes after forming the trench, depositing an adhesion layer within the source/drain region along a sidewall surface of the trench. In various embodiments, and after depositing the adhesion layer, the method further includes epitaxially growing a continuous first source/drain layer over the adhesion layer along the sidewall surface of the trench.

Abstract: A semiconductor structure includes a first pair of source/drain features (S/D), a first stack of channel layers connected to the first pair of S/D, a second pair of S/D, and a second stack of channel layers connected to the second pair of S/D. The first pair of S/D each include a first epitaxial layer having a first dopant, a second epitaxial layer having a second dopant and disposed over the first epitaxial layer and connected to the first stack of channel layers, and a third epitaxial layer having a third dopant and disposed over the second epitaxial layer. The second pair of S/D each include a fourth epitaxial layer having a fourth dopant and connected to the second stack of channel layers, and a fifth epitaxial layer having a fifth dopant and disposed over the fourth epitaxial layer. The first dopant through the fourth dopant are of different species.

Abstract: A training system and a training method for a domain-specific data model are provided. The training method includes configuring a computing device to perform the following processes: generating, by a training set generation module, a training data set based on a domain knowledge graph; updating the data model based on the training data set; generating, by the training set generation module, training input text corresponding to the domain knowledge graph; inputting the training input text into the data model to obtain training output text; evaluating and generating a score by an evaluation module based on a correlation between the training output text and the domain knowledge graph; and adjusting, by a reinforcement learning module, parameters of the data model according to the score and an optimization goal of the reward model until the score meets a training completion condition, taking the data model as the domain-specific data model.

Abstract: A method according to the present disclosure includes receiving a structure. The structure includes a substrate, a first fin-shaped structure, a second fin-shaped structure, and a third fin-shaped structure disposed over the substrate, and a first isolation feature between the first fin-shaped structure and the second fin-shaped structure and a second isolation feature between the second fin-shaped structure and the third fin-shaped structure. The method further includes depositing a first dielectric layer over the first isolation feature and the second isolation feature, depositing a second dielectric layer over the first dielectric layer and the first isolation feature, but not over the second isolation feature, performing a first selective etching process to the first dielectric layer and the second dielectric layer, and performing a second selective etching process to the first dielectric layer over the second isolation feature.

Abstract: A chip verification method based on SPI includes steps of (a) obtaining parameter information of a module of a chip under test and a corresponding register; (b) writing a test case for the module; (c) driving the parameter information in the test case onto an SPI bus at a top level of the chip under test through a driver, accessing a register of the module; (d) verifying the access result to determine whether the register is normal; (e) repeating steps (b) to (d) until all registers of the current module have been verified; and (f) repeating steps (a) to (e) until all modules of the chip have been verified. The method does not require using the chip bus to access its registers, thus alleviating the burden on the chip bus, reducing the probability of access errors, and improving the efficiency of chip verification.

Abstract: Well pick-up (WPU) regions are disclosed herein for improving performance of memory arrays, such as static random access memory arrays. An exemplary semiconductor device includes a circuit region, a first WPU region, second WPU region, a first well of a first conductivity type, and a second well of a second conductivity type. The circuit region, the first WPU region, and the second WPU region are arranged along a first direction in sequence. The first well has a first portion disposed in the circuit region and a second portion disposed in the first WPU region. The second well has a first portion disposed in the circuit region, a second portion disposed in the first WPU region, and a third potion disposed in the second WPU region. Measured along the first direction a width of the first WPU region is less than a width of the second WPU region.

Abstract: Various embodiments of the present disclosure are directed towards a method for forming an integrated chip. The method includes forming an epitaxial structure having a first doping type over a first portion of a semiconductor substrate. A second portion of the semiconductor substrate is formed over the epitaxial structure and the first portion of the semiconductor substrate. A first doped region having the first doping type is formed in the second portion of the semiconductor substrate and directly over the epitaxial structure. A second doped region having a second doping type opposite the first doping type is formed in the second portion of the semiconductor substrate, where the second doped region is formed on a side of the epitaxial structure. A plurality of fins of the semiconductor substrate are formed by selectively removing portions of the second portion of the semiconductor substrate.

Abstract: A semiconductor device includes a first memory cell and a dummy region adjacent to the first memory cell. The first memory cell includes a first transistor. The dummy region includes a cut-off transistor. The cut-off transistor has a first terminal electrically coupled to a second terminal of the first transistor. The cut-off transistor has a third terminal electrically coupled to ground.

Abstract: A method includes providing a substrate having a first region and a second region, forming a fin protruding from the first region, where the fin includes a first SiGe layer and a stack alternating Si layers and second SiGe layers disposed over the first SiGe layer and the first SiGe layer has a first concentration of Ge and each of the second SiGe layers has a second concentration of Ge that is greater than the first concentration, recessing the fin to form an S/D recess, recessing the first SiGe layer and the second SiGe layers exposed in the S/D recess, where the second SiGe layers are recessed more than the first SiGe layer, forming an S/D feature in the S/D recess, removing the recessed first SiGe layer and the second SiGe layers to form openings, and forming a metal gate structure over the fin and in the openings.

Abstract: Semiconductor devices having improved source/drain features and methods for fabricating such are disclosed herein. An exemplary device includes a semiconductor layer stack disposed over a mesa structure of a substrate. The device further includes a metal gate disposed over the semiconductor layer stack and an inner spacer disposed on the mesa structure of the substrate. The device further includes a first epitaxial source/drain feature and a second epitaxial source/drain feature where the semiconductor layer stack is disposed between the first epitaxial source/drain feature and the second epitaxial source/drain feature. The device further includes a void disposed between the inner spacer and the first epitaxial source/drain feature.

Abstract: SRAM designs based on GAA transistors are disclosed that provide flexibility for increasing channel widths of transistors at scaled IC technology nodes and relax limits on SRAM performance optimization imposed by FinFET-based SRAMs. GAA-based SRAM cells described have active region layouts with active regions shared by pull-down GAA transistors and pass-gate GAA transistors. A width of shared active regions that correspond with the pull-down GAA transistors are enlarged with respect to widths of the shared active regions that correspond with the pass-gate GAA transistors. A ratio of the widths is tuned to obtain ratios of pull-down transistor effective channel width to pass-gate effective channel width greater than 1, increase an on-current of pull-down GAA transistors relative to an on-current of pass-gate GAA transistors, decrease a threshold voltage of pull-down GAA transistors relative to a threshold voltage of pass-gate GAA transistors, and/or increases a ? ratio of an SRAM cell.

Abstract: Semiconductor device and the manufacturing method thereof are disclosed. An exemplary semiconductor device comprises a first semiconductor stack and a second semiconductor stack over a substrate, wherein each of the first and second semiconductor stacks includes semiconductor layers stacked up and separated from each other; a dummy spacer between the first and second semiconductor stacks, wherein the dummy spacer contacts a first sidewall of each semiconductor layer of the first and second semiconductor stacks; and a gate structure wrapping a second sidewall, a top surface, and a bottom surface of each semiconductor layer of the first and second semiconductor stacks.

Abstract: A method comprises forming a first fin including alternating first channel layers and first sacrificial layers and a second fin including alternating second channel layers and second sacrificial layers, forming a capping layer over the first and the second fin, forming a dummy gate stack over the capping layer, forming source/drain (S/D) features in the first and the second fin, removing the dummy gate stack to form a gate trench, removing the first sacrificial layers and the capping layer over the first fin to form first gaps, removing the capping layer over the second fin and portions of the second sacrificial layers to from second gaps, where remaining portions of the second sacrificial layers and the capping layers form a threshold voltage (Vt) modulation layer, and forming a metal gate stack in the gate trench, the first gaps, and the second gaps.

Abstract: A memory device includes a substrate, first semiconductor fin, second semiconductor fin, first gate structure, second gate structure, first gate spacer, and a second gate spacer. The first gate structure crosses the first semiconductor fin. The second gate structure crosses the second semiconductor fin, the first gate structure extending continuously from the second gate structure, in which in a top view of the memory device, a width of the first gate structure is greater than a width of the second gate structure. The first gate spacer is on a sidewall of the first gate structure. The second gate spacer extends continuously from the first gate spacer and on a sidewall of the second gate structure, in which in the top view of the memory device, a width of the first gate spacer is less than a width of the second gate spacer.