The present disclosure provides, in accordance with some illustrative embodiments, a semiconductor device structure including a hybrid substrate comprising an SOI region and a bulk region, the SOI region comprising an active semiconductor layer, a substrate material, and a buried insulating material interposed between the active semiconductor layer and the substrate material, and the bulk region being provided by the substrate material, an insulating structure formed in the hybrid substrate, the insulating structure separating the bulk region and the SOI region, and a gate electrode formed in the bulk region, wherein the insulating structure is in contact with two opposing sidewalls of the gate electrode.

A semiconductor device includes a semiconductor substrate, multiple memory cells on the semiconductor substrate arranged along a first dimension and along a second dimension that is orthogonal to the first dimension, in which each memory cell of the multiple memory cells includes a channel region in the semiconductor substrate, a tunnel dielectric layer on the channel region, and a first electrode layer on the tunnel dielectric layer. Along the first dimension, the channel region of each memory cell of the multiple memory cells is separated from the channel region of an adjacent memory cell of the multiple memory cells by a corresponding first air gap, each first air gap extending from below an upper surface of the semiconductor substrate up to an inter-electrode dielectric layer.

A power management circuit that has multiple sets of circuits to provide certain same power management functionalities in different power modes, such as voltage, current and temperature sensing and/or measuring, generating of reference states or biases to effectuate circuit protection in various conditions, such as under voltages, over voltages, etc. One set of circuits is configured to operate during a normal mode and is optimized for performance, speed and/or accuracy. Another set of circuits is configured to operate during a sleep mode and is optimized for reduced power consumption where the performance, speed and/or accuracy may be inferior to the circuits for the normal mode but the functionality is maintained within the low power consumption constraint.

Provided is a stable manufacturing method for a semiconductor device. In the manufacturing method for a semiconductor device, first, fins with an equal width are formed in each of a memory cell portion and a logic portion of a semiconductor substrate. Then, the fins in the logic portion are etched with the fins in the memory cell covered with a mask film, thereby fabricating fins in the logic portion, each of which is narrower than the fin formed in the memory cell portion.

A memory device and a method for forming the same are provided. The method includes forming a plurality of gate structures on a substrate, forming a first spacer on opposite sides of the gate structures, filling a dielectric layer between adjacent first spacers, forming a metal silicide layer on the gate structures, conformally forming a spacer material layer over the metal silicide layer, the first spacer layer and the dielectric layer, and performing an etch back process on the spacer material layer to form a second spacer on opposite sides of the metal silicide layer.

A semiconductor device is disclosed that comprises a first non-volatile memory cell, a second non-volatile memory cell, an active region between the first and second memory cells, and an electrically conductive contact touching the active region, wherein the contact has a horizontal cross-section that is at least five percent smaller in a first dimension than in a second dimension.

A semiconductor memory with a U-shaped channel comprises: a U-shaped channel region arranged in a semiconductor substrate, a source region, a drain region, a first layer of insulation film arranged on the U-shaped channel region, a floating gate provided with a notch, a second layer of insulation film, a control gate, a p-n junction diode arranged between the floating gate and the drain region, and a gate controlled diode formed by the control gate, the second layer of insulation film, and the p-n junction diode and using the control gate as a gate. Under the precondition of not increasing the manufacturing cost and difficulty of the semiconductor memory with a U-shaped channel and not affecting the performance of the semiconductor memory with a U-shaped channel, the dimension of a semiconductor storage device is further reduced and the chip density is increased by arranging the notch in the floating gate.

A memory device includes a capacitor, a tunneling-enhanced device, and a transistor. In accordance with an embodiment, capacitor has first and second electrodes wherein the first electrode of the capacitor serves as a control gate of the memory device. The tunneling-enhanced device has a first electrode and a second electrode, wherein the first electrode of the second capacitor serves as an erase gate of the memory device and the second electrode of the tunneling-enhanced device is coupled to the second electrode of the capacitor to form a floating gate. The transistor has a control electrode and a pair of current carrying electrodes, wherein the control electrode of the transistor is directly coupled to the floating gate. In accordance with another embodiment, a method for manufacturing the memory device includes a method for manufacturing the memory device.

A method of manufacturing an embedded flash memory device is provided. A pair of gate stacks are formed spaced over a semiconductor substrate, and including floating gates and control gates over the floating gates. A common gate layer is formed over the gate stacks and the semiconductor substrate, and lining sidewalls of the gate stacks. A first etch is performed into the common gate layer to recess an upper surface of the common gate layer to below upper surfaces respectively of the gate stacks, and to form an erase gate between the gate stacks. Hard masks are respectively formed over the erase gate, a word line region of the common gate layer, and a logic gate region of the common gate layer. A second etch is performed into the common gate layer with the hard masks in place to concurrently form a word line and a logic gate.

A method is provided that includes forming a vertical bit line disposed in a first direction above a substrate, forming a multi-layer word line disposed in a second direction above the substrate, the second direction perpendicular to the first direction, and forming a memory cell including a nonvolatile memory material at an intersection of the vertical bit line and the multi-layer word line. The multi-layer word line includes a first conductive material layer and a second conductive material layer disposed above the first conductive material layer. The memory cell includes a working cell area encompassed by an intersection of the first conductive material layer and the nonvolatile memory material.