A non-volatile memory having memory cells is provided. The memory cells include stack structures, floating gates, tunneling dielectric layers, erase gate dielectric layers, auxiliary gate dielectric layers, source regions, drain regions, control gates and inter-gate dielectric layers. The stacked structures include gate dielectric layers, auxiliary gates, insulating layers and erase gates. The floating gates are disposed on sidewalls on a first side of the stacked structures. The tunneling dielectric layers are disposed under the floating gates. The erase gate dielectric layers are disposed between the erase gates and floating gates. The auxiliary gate dielectric layers are disposed between the auxiliary gates and the floating gates. The source and drain regions are separately disposed on sides of the stack structures and the floating gates. The control gates are disposed on the source regions and the floating gates. The inter-gate dielectric layers are disposed between the control gates and the floating gates.

A semiconductor memory device includes a substrate having a first active area and a second active area in proximity to the first active area. A trench isolation region is between the first active area and the second active area. A source line region is disposed in the first active area and adjacent to the trench isolation region. An erase gate is disposed on the source line region. A floating gate is disposed on a first side of the erase gate. A first control gate is disposed on the floating gate. A first word line is disposed adjacent to the floating gate and the first control gate and insulated therefrom. A second control gate is disposed on a second side of the erase gate and directly on the trench isolation region. A second word line is disposed adjacent to the second control gate and insulated therefrom.

A program method of a nonvolatile memory device that performs a plurality of program loops is provided. At least one of the plurality of program loops includes dividing a channel of a selected cell string into a first side channel and a second side channel during a first interval and a second interval, turning off a string selection transistor of the selected cell string by applying a string select line voltage of a first level during the first interval, and boosting a first voltage of the first side channel and a second voltage of the second side channel, and turning on the string selection transistor by applying the string select line voltage of a second level different from the first level during the second interval, and performing a hot carrier injection (HCI) program operation on a selected memory cell corresponding to the first side channel or the second side channel.

A nonvolatile memory device is provided. The nonvolatile memory device comprises a floating gate arranged over a first active region, whereby the first active region is in an active layer of a substrate. A metal-insulator-metal (MIM) capacitor may be provided laterally adjacent to the floating gate, whereby a portion of the metal-insulator-metal capacitor is in the active layer. A contact pillar may connect a first electrode of the metal-insulator-metal capacitor to the floating gate.

Numerous embodiments are disclosed for a high voltage generation algorithm and system for generating high voltages necessary for a particular programming operation in analog neural memory used in a deep learning artificial neural network. In one example, a method for programming a plurality of non-volatile memory cells in an array of non-volatile memory cells, comprises generating a high voltage, and programming a plurality of non-volatile memory cells in an array using the high voltage when a programming enable signal is asserted and providing a feedback loop to maintain the high voltage while programming the plurality of non-volatile memory cells.

A method for programming a memory. The method includes providing a memory structure with a floating gate, and grounding a source of the memory structure; applying voltages to a drain and a bulk, forming an electric field, generating electron-hole pairs, and generating primary electrons, wherein the voltage applied to the bulk is lower than the voltage applied to the drain; making holes accelerate downward under the action of the electric field and collide with the bulk in the memory structure within a predetermined time to generate secondary electrons; applying voltages to a gate and the bulk respectively, where the voltage applied to the bulk is lower than the voltage applied to the gate, to enable the secondary electrons to generate tertiary electrons under the action of an electric field in a vertical direction, and the tertiary electrons are injected into the floating gate to complete a programming operation.

In one aspect, a flash memory cell includes a well having a first-type dopant, a source having a second-type dopant and formed within the well, a drain having the second-type dopant and formed within the well, a floating gate above the well, a control gate above the floating gate, an oxide compound disposed between the floating gate and the control gate, and a tunnel oxide disposed between the floating gate and the well. The flash memory cell is configured, in one of a program mode or an erase mode, to move an electron from the source to the floating gate. The flash memory cell is configured, in the other one of the program or the erase mode, to move an electron is from the floating gate to the drain.

A memory cell array with memory cells arranged in rows and columns, first sub source lines each connecting together the source regions in one of the rows and in a first plurality of the columns, second sub source lines each connecting together the source regions in one of the rows and in a second plurality of the columns, a first and second erase gate lines each connecting together all of the erase gates in the first and second plurality of the columns respectively, first select transistors each connected between one of first sub source lines and one of a plurality of source lines, second select transistors each connected between one of second sub source lines and one of the source lines, first select transistor line connected to gates of the first select transistors, and a second select transistor line connected to gates of the second select transistors.

An erasable programmable non-volatile memory includes a first-type well region, three doped regions, two gate structures, a blocking layer and an erase line. The first doped region is connected with a source line. The third doped region is connected with a bit line. The first gate structure is spanned over an area between the first doped region and the second doped region. A first polysilicon gate of the first gate structure is connected with a select gate line. The second gate structure is spanned over an area between the second doped region and the third doped region. The second gate structure includes a floating gate and the floating gate is covered by the blocking layer. The erase line is contacted with the blocking layer. The erase line is located above an edge or a corner of the floating gate.

An integrated circuit includes two different types of embedded memories, with cells that have different retention characteristics, and situated in different areas of the substrate. In some applications the cells are both non-volatile memories sharing a common gate layer but with different oxide layers, different thicknesses, etc. The first type of cell is a conventional flash cell which can be part of a logic/memory region, while the second type of cell uses capacitive coupling and can be located in a high voltage region. Because of their common features, the need for additional masks, manufacturing steps, etc. can be mitigated.