A resistive memory device includes a plurality of first conductive lines in a first area and a second area on a substrate, a plurality of second conductive lines in the first area and the second area, the plurality of second conductive lines being apart from the plurality of first conductive lines in a vertical direction, and a plurality of memory cells connected to the first and second conductive lines at a plurality of intersections between the plurality of first and second conductive lines in the first area and the second area. The plurality of memory cells include an active memory cell in the first area and a dummy memory cell in the second area. The active memory cell including a first resistive memory pattern having a first width and the dummy memory cell including a second resistive memory pattern having a second width greater than the first width.

An artificial neuron semiconductor device having a three-dimensional structure includes a first electrode to which a clock signal is applied, a second electrode in which an output signal is generated, an insulation column, a plurality of electrode layers for receiving an electrical signal from at least one synapse circuit, and a phase change layer which is divided into at least two parts by the insulation column and is in contact with at least two side surfaces of the insulation column, and the phase change layer is phase-changed by the plurality of electrode layers.

Some embodiments include apparatus and methods having a memory device with diodes coupled to memory elements. Each diode may be formed in a recess of the memory device. The recess may have a polygonal sidewall. The diode may include a first material of a first conductivity type (e.g., n-type) and a second material of a second conductive type (e.g., p-type) formed within the recess.

A phase change memory cell has first and second electrodes having phase change material there-between. The phase change memory cell is devoid of heater material as part of either of the first and second electrodes and being devoid of heater material between either of the first and second electrodes and the phase change material. A method of forming a memory cell having first and second electrodes having phase change material there-between includes lining elevationally inner sidewalls of an opening with conductive material to comprise the first electrode of the memory cell. Elevationally outer sidewalls of the opening are lined with dielectric material. Phase change material is formed in the opening laterally inward of and electrically coupled to the conductive material in the opening. Conductive second electrode material is formed that is electrically coupled to the phase change material. Other implementations are disclosed.

The present disclosure includes memory cell array structures and methods of forming the same. One such array includes a stack structure comprising a memory cell between a first conductive material and a second conductive material. The memory cell can include a select element and a memory element. The array can also include an electrically inactive stack structure located at an edge of the stack structure.

A three-dimensional resistive memory device includes an alternating stack of electrically conductive layers and insulating layers. Resistive memory elements are provided between the electrically conductive layers and a semiconductor local bit line. The semiconductor local bit line includes a heterostructure of an inner semiconductor material layer having an inner-material band gap and an outer semiconductor material layer having an outer-material band gap that is narrower than the inner-material band. A gate dielectric is located between a gate electrode and the inner semiconductor material layer.

According to one embodiment, a semiconductor memory device includes a semiconductor layer, a gate electrode, a metal containing portion, and an insulating portion. The semiconductor layer includes a first region and a second region. The second region has at least one of a region being amorphous or a region having a crystallinity lower than a crystallinity of the first region. The gate electrode is apart from the first region in a first direction. The first direction crosses a second direction connecting the first region and the second region. The metal containing portion is apart from the second region in the first direction. At least a part of the metal containing portion overlaps the gate electrode in the second direction. The insulating portion is provided between the gate electrode and the first region and between the metal containing portion and the second region.

The present invention is a method of incorporating a non-volatile memory into a CMOS process that requires four or fewer masks and limited additional processing steps. The present invention is an epi-silicon or poly-silicon process sequence that is introduced into a standard CMOS process (i) after the MOS transistors' gate oxide is formed and the gate poly-silicon is deposited (thereby protecting the delicate surface areas of the MOS transistors) and (ii) before the salicided contacts to those MOS transistors are formed (thereby performing any newly introduced steps having an elevated temperature, such as any epi-silicon or poly-silicon deposition for the formation of diodes, prior to the formation of that salicide). A 4F.sup.2 memory array is achieved with a diode matrix wherein the diodes are formed in the vertical orientation.

Providing a high-density two-terminal memory architecture(s) having performance benefits of two-terminal memory and relatively low fabrication cost, is described herein. By way of example, the two-terminal memory architecture(s) can be constructed on a substrate, in various embodiments, and comprise two-terminal memory cells formed within conductive layer recess structures of the memory architecture. In one embodiment, a conductive layer recess can be created as a horizontal etch in conjunction with a vertical via etch. In another embodiment, the conductive layer recess can be patterned for respective conductive layers of the two-terminal memory architecture.

Methods, systems, and devices for memory arrays that use a conductive hard mask during formation and, in some cases, operation are described. A hard mask may be used to define features or components during the numerous material formation and removal steps used to create memory cells within a memory array. The hard mask may be an electrically conductive material, some or all of which may be retained during formation. A conductive line may be connected to each memory cell, and because the hard mask used in forming the cell may be conductive, the cell may be operable even if portions of the hard mask remain after formation.