Abstract: A three-dimensional (3D) memory device is disclosed. The 3D memory device comprises an alternating layer stack on a substrate, a plurality of channel holes penetrating the alternating layer stack, a channel structure in each channel hole, and a top selective gate cut structure having a laminated structure and located between two rows of channel structures.

Abstract: A method for forming a three-dimensional memory device can include forming a staircase structure. An alternating layer stack is disposed and etched to form steps. A continuous layer disposed on the staircase structure continuously extends over the steps. An insulating layer is disposed on the continuous layer and a slit is formed extending through the staircase structure. The slit exposes sidewalls of the continuous layer and the steps. The sacrificial layer is removed and a cavity is formed in place of the continuous layer. An etch stop layer is disposed in the cavity and continuously extends over the steps. Openings are formed through the insulating layer and expose a portion of a lateral surface of the etch stop layer. The openings are extended through the etch stop layer to expose a lateral surface of each step of the steps. Contacts are formed in the openings.

Abstract: A semiconductor device semiconductor device includes a stack having a first surface and a second surface opposing the first surface. The stack can include word line layers and insulating layers alternating with the word line layers between the first surface and the second surface. The stack can further include a process stop layer between the lower most insulating layer and the second surface. The stack can extend along an X-Y plane having an X direction and a Y direction perpendicular. The semiconductor device can further include a slit structure crossing the stack between the first surface and the second surface in Z direction. In a cross-section perpendicular to the Y direction, distances between the slit structure and the process stop layer at two sides of the slit structure are each larger than distances at either side of the slit structure between the word line layers and the slit structure.

Abstract: A side impact protection mechanism is provided and is disposed on a lateral wing of any component of a child carrier and includes a side impact protection block, a locking component and an operating component. The side impact protection block is switchable between a folded position and an unfolded position, so that the side impact protection block is closely fitted with or at least partially protrudes from the lateral wing. The locking component is switchable between a locking position and a releasing position, so that the side impact protection block is restrained from leaving from the unfolded position or allowed to pivot from the unfolded position to the folded position. The operating component can drive the locking component to move toward the releasing position.

Abstract: A three-dimensional (3D) memory device is disclosed. The 3D memory device comprises an alternating layer stack on a substrate, and a top selective gate cut structure having a laminated structure embedded in an upper portion of the alternating layer stack and extending along a lateral direction. The laminated structure of the top selective gate cut structure comprises a dielectric filling wall and a dummy channel and a dummy functional layer on both sides of the dielectric filling wall.

Abstract: The present disclosure discloses a child safety seat, which includes: a base, a rear portion of which is formed with a top rod protruding upwardly; a seat rotatably fixed on the base; and a tether telescopically connected to the base, wherein, the tether can be automatically retracted when it is extended relative to the base. The child safety seat of the present disclosure has an advantage of convenient connection and is not easy to misuse.

Abstract: Disclosed are a fault early-warning method and a fault early-warning system applied to a gas turbine unit, and an apparatus. The method includes: calculating, by means of a mechanism model, predicted data of prediction parameters in a gas turbine unit, and performing data comparison on the predicted data and real data of the prediction parameters, so as to obtain an error matrix; constructing several Kriging primary functions according to the mechanism model; screening an optimal Kriging primary function from the several Kriging primary functions according to the error matrix, and performing error compensation on the mechanism model by using the optimal Kriging primary function; and performing fault early warning on the gas turbine unit by means of the mechanism model after error compensation. By means of the present application, the problem of a large number of fault false alarms in fault early warning of an existing gas turbine is solved.

Abstract: A three-dimensional (3D) memory device includes a plurality of memory planes and a separation block. Each memory plane includes a plurality of memory blocks. Each memory block includes a memory stack including interleaved conductive layers and first dielectric layers, and a plurality of channel structures each extending through the memory stack. The separation block extending laterally to separate each two adjacent memory planes. Each separation block includes a dielectric stack including interleaved second dielectric layers and the first dielectric layers. The first dielectric layers extend across the memory blocks and the separation block, and the second dielectric layers separate the conductive layers of two adjacent memory blocks.

Abstract: Systems and methods for providing a neural network system for time series forecasting are described. A time series dataset that includes datapoints at a plurality of timestamps in an observed space is received. The neural network system is trained using the time series dataset. The training the neural network includes: generating, using an encoder of the neural network system, one or more estimated latent variables of a latent space for the time series dataset; generating, using an auxiliary predictor of the neural network system, a first latent-space prediction result based on the one or more estimated latent variables; transforming, using a decoder of the neural network system, the first latent-space prediction result to a first observed-space prediction result; and updating parameters of the neural network system based on a loss based on the first observed-space prediction result.

Abstract: Three-dimensional (3D) memory devices and methods for forming the same are disclosed. The 3D memory device includes a doped semiconductor layer, a source select gate line disposed on the doped semiconductor layer, a stack structure including interleaved conductive layers and dielectric layers formed on the source select gate line, and a channel structure extending through the stack structure and the source select gate line and in contact with the doped semiconductor layer. The channel structure includes a semiconductor channel and a memory film. The source select gate line is in contact with the semiconductor channel.

Abstract: Embodiments of 3D memory devices having a concentric staircase structure and methods for forming the same are disclosed. In an example, a 3D memory device includes a memory array structure and a concentric staircase structure in an intermediate of the memory array structure. The concentric staircase structure includes a plurality of concentric zones in a radial direction in a plan view. Each of the plurality of concentric zones includes a plurality of stairs in a tangential direction in the plan view.

Abstract: Disclosed are a fault early-warning method and a fault early-warning system applied to a gas turbine unit, and an apparatus. The method includes: calculating, by means of a mechanism model, predicted data of prediction parameters in a gas turbine unit, and performing data comparison on the predicted data and real data of the prediction parameters, so as to obtain an error matrix; constructing several Kriging primary functions according to the mechanism model; screening an optimal Kriging primary function from the several Kriging primary functions according to the error matrix, and performing error compensation on the mechanism model by using the optimal Kriging primary function; and performing fault early warning on the gas turbine unit by means of the mechanism model after error compensation. By means of the present application, the problem of a large number of fault false alarms in fault early warning of an existing gas turbine is solved.

Abstract: A memory device includes a substrate; and a stack structure, including alternately arranged first dielectric layers and electrode layers. In a first lateral direction, the memory device includes an intermediate region and array regions. In a second lateral direction, the stack structure includes a first block and a second block, each including a wall-structure region. In the intermediate region, wall-structure regions of the first block and the second block are separated by a staircase structure. The memory device further includes a beam structure, located in the intermediate region and including at least a plurality of discrete first beam structures, each extending along the second lateral direction and connecting the wall-structure regions of the first block and the second block; and a plurality of second dielectric layers, located in the beam structure. In the first beam structures, the second dielectric layers is alternated with the first dielectric layers.

Abstract: Systems and methods for providing a neural network system for time series forecasting are described. A time series dataset that includes datapoints at a plurality of timestamps in an observed space is received. A first state-space model of a dynamical system underlying the time series dataset is provided. The first state-space model includes a non-parametric latent transition model. One or more latent variables of a latent space for the time series dataset are determined using the neural network system based on the first state-space model. A first prediction result for the time series dataset is provided by the neural network system based on the estimated latent variables.

Abstract: A screen projection display method and apparatus, a mobile terminal, a storage medium, and a program product. The method comprises: when the interface content of a first application program is displayed on a screen projection display device, receiving a first instruction (201); in response to the first instruction, displaying an application selection interface, the application selection interface comprising an option of at least one second application program (202); receiving a second instruction for a target second application program among the at least one second application (203); and in response to the second instruction, projecting the interface content of the target second application program to the screen projection display device for display (204).

Abstract: Embodiments of three-dimensional memory devices and fabricating methods thereof are disclosed. One disclosed method for forming a memory structure comprises: forming a bottom conductive layer on a substrate; forming a dielectric stack on the bottom conductive layer, the dielectric stack comprising a plurality of alternatively arranged first dielectric layers and second dielectric layers; forming an opening penetrating the dielectric stack and exposing the bottom conductive layer; forming a cap layer on a bottom of the opening; forming a cylindrical body and a top contact on the cap layer and in the opening; and replacing the plurality of second dielectric layers with conductive layers.

Abstract: Embodiments of three-dimensional memory devices are disclosed. A disclosed memory structure comprises a memory cell comprising: a cylindrical body having a cylindrical shape, an insulating layer surrounding the cylindrical body, a word line contact coupled to a word line and surrounding a first portion of the insulating layer, and multiple plate line contact segments coupled to multiple plate lines respectively and surrounding a second portion of the insulating layer. The memory structure further comprises a bit line contact coupled to a bit line and coupled to a first end of the cylindrical body, a source line contact coupled to a source line, and a source cap coupled between the source line contact and a second end of the cylindrical body to increase a distance between the source line contact and the plate line contact segments.

Abstract: Embodiments of three-dimensional memory devices and fabricating methods thereof are disclosed. One disclosed method for forming a memory structure comprises: forming a bottom conductive layer on a substrate; forming a memory stack on the bottom conductive layer, the memory stack comprising a plurality of alternatively arranged dielectric layers and conductive layers; forming an opening penetrating the memory stack and exposing the bottom conductive layer; forming a cap layer on a bottom of the opening; forming a cylindrical body and a top contact on the cap layer and in the opening; and forming a plurality of interconnection structures to electrically connect the bottom conductive layer, the plurality of conductive layers of the memory stack, and the top contact.

Abstract: A permeable pavement system including a permeable pavement composition and a related method are provided. The permeable pavement system includes a first layer of a permeable pavement composition including a quantity of a first permeable pavement material and a quantity of cured carbon fiber composite material (CCFCM) incorporated therewith, the first layer defining a first surface; and a second layer of a second permeable pavement material deposited over a substantial entirety of and covering the first surface of the first layer of the permeable pavement composition, wherein the first layer interfaces with the second layer to at least strengthen the permeable pavement system.