A semiconductor memory device includes a first semiconductor layer that includes a first part extending in a first direction, a second part extending in the first direction, and a third part connected to the first and second parts. When a cross-sectional surface extending in second and third directions and including the third part is defined as a first cross-sectional surface, the third part has one side and the other side of an imaginary center line in the third direction in the first cross-sectional surface defined as first and second regions, the third part has maximum widths in the second direction in the first and second regions defined as first and second widths, and the third part has a width in the second direction on the imaginary center line defined as a third width, the third width is smaller than the first and second widths.

A neuromorphic device for the analog computation of a linear combination of input signals, for use, for example, in an artificial neuron. The neuromorphic device provides non-volatile programming of the weights, and fast evaluation and programming, and is suitable for fabrication at high density as part of a plurality of neuromorphic devices. The neuromorphic device is implemented as a vertical stack of flash-like cells with a common control gate contact and individually contacted source-drain (SD) regions. The vertical stacking of the cells enables efficient use of layout resources.

A neuromorphic device for the analog computation of a linear combination of input signals, for use, for example, in an artificial neuron. The neuromorphic device provides non-volatile programming of the weights, and fast evaluation and programming, and is suitable for fabrication at high density as part of a plurality of neuromorphic devices. The neuromorphic device is implemented as a vertical stack of flash-like cells with a common control gate contact and individually contacted source-drain (SD) regions. The vertical stacking of the cells enables efficient use of layout resources.

Embodiments of three-dimensional (3D) memory devices and methods for forming the same are disclosed. In an example, a method for forming a 3D memory device comprises forming a channel structure extending vertically through a memory stack into a semiconductor layer on a substrate. The memory stack comprises interleaved stack conductive layers and stack dielectric layers. The method further comprises forming an insulating structure in an opening extending vertically through the memory stack and at a distance away from the channel structure, and comprising a dielectric layer doped with at least one of hydrogen or an isotope of hydrogen.

Embodiments of three-dimensional (3D) memory devices and methods for forming the same are disclosed. In an example, a method for forming a 3D memory device comprises forming a channel structure extending vertically through a memory stack into a semiconductor layer on a substrate. The memory stack comprises interleaved stack conductive layers and stack dielectric layers. The method further comprises forming an insulating structure in an opening extending vertically through the memory stack and at a distance away from the channel structure, and comprising a dielectric layer doped with at least one of hydrogen or an isotope of hydrogen.

Provided is a method of manufacturing a memory device and a patterning method. The patterning method includes following steps. A control structure including stop layers and oxide layers stacked alternately, a hard mask layer, and a mask pattern are sequentially formed on a target layer. A photoresist layer is formed in the mask pattern on the hard mask layer. A portion of the hard mask layer and a portion of the control structure are removed to form first openings by using the photoresist layer and the mask pattern as a mask. The photoresist layer and the hard mask layer are removed to form a second opening having a bottom surface higher than that of the first openings. At least one etching process is performed so that the first and second openings extend into and divide the control structure and the target layer into stack structures.

Provided is a method of manufacturing a memory device and a patterning method. The patterning method includes following steps. A control structure including stop layers and oxide layers stacked alternately, a hard mask layer, and a mask pattern are sequentially formed on a target layer. A photoresist layer is formed in the mask pattern on the hard mask layer. A portion of the hard mask layer and a portion of the control structure are removed to form first openings by using the photoresist layer and the mask pattern as a mask. The photoresist layer and the hard mask layer are removed to form a second opening having a bottom surface higher than that of the first openings. At least one etching process is performed so that the first and second openings extend into and divide the control structure and the target layer into stack structures.

Provided are a memory device and a method of manufacturing the same. The memory device includes: a stack structure; a first source/drain region and a second source/drain region located in a substrate beside the stack structure; a first self-aligned contact connected to the first source/drain region; a second self-aligned contact connected to the second source/drain region; a first liner structure located between the first self-aligned contact and a first sidewall of the stack structure; and a second liner structure located between the second self-aligned contact and a second sidewall of the stack structure. The first liner structure and the second liner structure are not connected and do not cover the stack structure.

Provided are a memory device and a method of manufacturing the same. The memory device includes: a stack structure; a first source/drain region and a second source/drain region located in a substrate beside the stack structure; a first self-aligned contact connected to the first source/drain region; a second self-aligned contact connected to the second source/drain region; a first liner structure located between the first self-aligned contact and a first sidewall of the stack structure; and a second liner structure located between the second self-aligned contact and a second sidewall of the stack structure. The first liner structure and the second liner structure are not connected and do not cover the stack structure.

A nonvolatile memory (“NVM”) bitcell with one or more active regions capacitively coupled to the floating gate but that are separated from both the source and the drain. The inclusion of capacitors separated from the source and drain allows for improved control over the voltage of the floating gate. This in turn allows CHEI (or IHEI) to be performed with much higher efficiency than in existing bitcells, thereby the need for a charge pump to provide current to the bitcell, ultimately decreasing the total size of the bitcell. The bitcells may be constructed in pairs, further reducing the space requirements of the each bitcell, thereby mitigating the space requirements of the separate capacitor/s. The bitcell may also be operated by CHEI (or IHEI) and separately by BTBT depending upon the voltages applied at the source, drain, and capacitor/s.