Abstract: A method of fabricating a magnetic storage device includes depositing a first conductive material. The method further includes electrically isolating distinct instances of the first conductive material to form a first wire extending along a first direction. The method further includes depositing, on the distinct instances of the first conductive material, a set of device layers. The method further includes electrically isolating distinct instances of the device layers to form spin orbit torque magnetic random access memory (SOT-MRAM) devices positioned on distinct instances of the first conductive material. The method further includes depositing, on the distinct instances of the device layers, a layer of a second conductive material and electrically isolating a plurality of distinct instances of the layer of the second conductive material to form a plurality of second wires extending along a second direction. The second direction is different from the first direction.

Abstract: A method for manufacturing a magnetic memory element structure using a Ru hard mask and a post pillar thermal annealing process. A Ru hard mask is formed over a plurality of memory element layers and an ion milling is performed to transfer the image of the Ru hard mask onto the underlying memory element layers. A high-angle ion milling an be performed to remove any redeposited material from the sides of the memory element layers, and a non-magnetic, dielectric material can be deposited. A thermal annealing process can then be performed to repair any damage caused by the previously performed ion milling processes.

Abstract: According to one embodiment, a method includes forming a first insulative layer above a bottom surface of a groove and along inner sidewalls thereof, forming a source line layer within the groove of the substrate, forming a first dielectric layer on outer sides of a middle portion of the source line layer, forming a buffer layer on outer sides of the first dielectric layer, forming a gate terminal above the source line layer, forming a gate dielectric layer between the source line layer and the gate terminal and on outer sides of the lower portion of the gate terminal, forming a drain terminal including strained Si on outer sides of the first dielectric layer, and forming a relaxed buffer layer on outer sides of the upper portion of the source line layer and outer sides of the drain terminal, with the gate terminal extending beyond the relaxed buffer layer thickness.

Abstract: A method for manufacturing a magnetic memory element array that includes the use of a Ru hard mask layer and a diamond like carbon hard mask layer formed over the Ru hard mask layer. A plurality of magnetic memory element layers are deposited over a wafer and a Ru hard mask layer is deposited over the plurality of memory element layers. A layer of diamond like carbon is deposited over the Ru hard mask layer, and a photoresist mask is formed over the layer of diamond like carbon. A reactive ion etching is then performed to transfer the image of the photoresist mask onto the diamond like carbon mask, and an ion milling is performed to transfer the image of the patterned diamond like carbon mask onto the underlying Ru hard mask and memory element layers. The diamond like carbon mask can then be removed by reactive ion etching.

Abstract: An memory device comprising an array of memory cells wherein each memory cell includes a respective magnetic random access memory (MRAM) element, and a respective gating transistor. A plurality of bit lines are routed parallel to each other, wherein each bit line is associated with a respective memory cell of the array of memory cells. A common word line is coupled to gates of gating transistors of the array of memory cells. A common source line is coupled to sources of the gating transistors, wherein the common source line is routed perpendicular to the plurality of bit lines within the array of memory cells. A first circuit provides a first voltage on an addressed bit line of the plurality of bit lines during a write cycle, wherein the addressed bit line corresponds to an addressed memory cell.

Abstract: Embodiments of the present invention facilitate efficient and effective increased memory cell density configuration. In one embodiment, a semiconductor device comprises: a first pillar magnetic tunnel junction (pMTJ) memory cell that comprises a first pMTJ located in a first level in the semiconductor device; and a second pillar magnetic tunnel junction (pMTJ) memory cell that comprises a second pMTJ located in a second level in the semiconductor device, wherein the second pMTJ location with respect to the first pMTJ is coordinated to comply with a reference pitch for the memory cell. A reference pitch is associated a first switch coupled to the first pMTJ and the second pitch reference component is a second switch coupled to the second pMTJ. The first switch and second switch can be transistors. The reference pitch coordination facilitates reduced pitch between memory cells and increased information storage capacity of bits per memory device area.

Abstract: The various implementations described herein include magnetic memory devices and systems, and methods for propagating defects in the devices and systems. In one aspect, a magnetic memory device comprises a non-magnetic cylindrical core configured to receive a current, a plurality of magnetic layers surrounding the core, and a plurality of non-magnetic layers also surrounding the core. The magnetic layers and the non-magnetic layers are arranged in a stack coaxial with the core. Respective magnetic layers of the plurality of magnetic layers are separated by respective non-magnetic layers of the plurality of non-magnetic layers. The device further comprises an input terminal coupled to a first end of the core and a current source coupled to the input terminal. The current source is configured to supply current imparting a Spin Hall Effect (SHE) around the circumference of the core, and the SHE contributes to a magnetization of the magnetic layers.

Abstract: Semiconductor substrate adaptor configured to adapt a substrate of a first dimension to a second dimension, such that the substrate can be properly supported by a supporting mechanism (e.g., a wafer cassette) customized for substrates of the second dimension. The substrate adaptor may be made of quartz. The combination of the substrate adaptor and a substrate fitting therein causes no perturbation in various aspects of a semiconductor process. Therefore, the substrate adaptor conveniently enables a substrate of the first dimension to be processed in the same processing equipment and conditions as a substrate of the second dimension. A vertical substrate adaptor may have a semicircular body with a semicircular cutout for accommodating a wafer and can support a wafer vertically. A horizontal substrate adaptor may have a circular body with a circular cutout for accommodating an entire wafer and supporting the wafer horizontally.

Abstract: A magnetic memory element for using in magnetic random access memory. The magnetic memory element includes a novel exchange coupling layer for use in an antiferromagnetic structure for magnetically pinning a magnetic reference layer of the memory element. The exchange coupling layer is located between a first magnetic layer (reference layer) and a second magnetic layer (keeper layer). The exchange coupling layer includes a layer of Ru located between first and second layers of Ir. The Ir layers can be in contact with each of the first and second magnetic layers to provide an interfacial magnetic anisotropy, as well as providing RKKY exchange field. The Ru layer, provides an increased RKKY exchange field as a result of the high RKKY exchange coupling of Ru.

Abstract: A method for forming self aligned magnetic memory element pillars for Magnetic Random Access Memory. The method allows the magnetic memory element pillars to be arranged in staggered rows of memory elements at a pitch that is smaller than what is possible using photolithography alone. The method involves forming a spacer mask in the form of an array of connected rings arranged in a square pattern of non-staggered rows. A sacrificial mask material is deposited over the spacer mask and the spacer mask is then removed, leaving sacrificial mask material in the holes at the center of the rings and also in the spaces between the rings. A reactive ion processes is then performed to transfer the pattern of the sacrificial mask onto underlying hard mask layers. A material removal process can then be performed to define a plurality of memory element pillars.

Abstract: A magnetic memory array having a source-plane electrically connected with an array of channel selectors in two-dimensions. The array of channel selectors can be arranged in rows and columns with both the rows and columns being electrically connected with a source-plane. A memory element such as a two terminal resistive switching memory element can be electrically connected with each of the channel selectors. The source-plane can include a doped region formed in a surface of a semiconductor substrate and may also include an electrically conductive layer formed on the doped region. The use of such a planar, two-dimensional source-plane allows for greatly increased data density by eliminating the need to form separate source-line source lines for individual rows of channel selectors.

Abstract: The various implementations described herein include methods, devices, and systems for performing operations on memory devices. In one aspect, a memory device a magnetic memory component and a current selector component coupled to the magnetic memory component. The current selector component includes a first transistor having a first gate with a corresponding first threshold voltage. The first transistor comprises a charge storage layer configured to selectively store charge so as to adjust a current through the first transistor. The memory device further includes control circuitry configured to determine a bit error rate of the magnetic memory component and adjust a charge stored in the charge storage layer based on the determined bit error rate.

Abstract: A magnetic data recording system utilizing different magnetic memory element types to optimize competing performance parameters in a common memory chip. The memory system includes a first memory portion which can be a main memory and which includes magnetic memory elements of a first type, and a second memory region which can be a temporary memory region and which includes magnetic memory elements of a second type. A memory controller can be provided for controlling the input and retrieval of data to and from the first and second memory elements. The second, memory region can be a scratchpad memory or could also be cache type memory. The first type of magnetic memory elements can be designed for high data retention, whereas the second type of magnetic memory elements can be designed for fast write speed (low latency) and low write power consumption.

Abstract: A magnetic memory device that includes magnetic read elements and magnetic reference cells. The magnetic reference cells include magnetic tunnel junction elements having the same construction as the magnetic read elements. The reference cells produce a reference signal that can be compared with a read signal from the magnetic read element to determine whether the read element is in a high or low resistance state. During creation of the reference signal, the current passes in such a way so that reference cells are forced to be in the right state while causing no disturbance to the reference cell. The reference cell includes magnetic tunnel junction elements and also includes circuitry configured to produce a magnetic field that biases the magnetic tunnel junction elements of the reference cell into a desired magnetic state to ensure that the desired magnetic state of the reference cell magnetic tunnel junction elements is maintained.

Abstract: Embodiments of the present invention facilitate efficient and effective increased memory cell density configuration. In one embodiment, the method comprises: forming a first pitch reference component and a second pitch reference component; forming a first pillar magnetic tunnel junction (pMTJ) located in a first level and a second pMTJ located in a second level, wherein the location of the second pMTJ with respect to the first pMTJ is coordinated based upon a reference pitch distance between the first pitch reference component and first pitch reference component. In one exemplary implementation, the first pitch reference component is a first switch coupled to the first pMTJ and the second pitch reference component is a second switch coupled to the second pMTJ. The reference component size can be based upon a minimum lithographic processing dimension.

Abstract: A magnetic memory array having a source-plane electrically connected with an array of channel selectors in two-dimensions. The array of channel selectors can be arranged in rows and columns with both the rows and columns being electrically connected with a source-plane. A magnetic memory element such as a magnetic tunnel junction element can be electrically connected with each of the channel selectors. The source-plane can include a doped region formed in a surface of a semiconductor substrate and may also include an electrically conductive layer formed on the doped region. The use of such a planar, two-dimensional source-plane allows for greatly increased data density by eliminating the need to form separate source-line source lines for individual rows of channel selectors.

Abstract: A Magnetic Tunnel Junction (MTJ) device can include a free magnetic layer having a predetermined smoothness. An etching process for smoothing the free magnetic layer can be performed in-situ with various deposition processes after a high temperature annealing of the MTJ formation.

Abstract: A memory device comprises a memory bank comprising a plurality of memory addresses. The memory device further comprises a first level dynamic redundancy register comprising data storage elements and a pipeline bank coupled to the memory bank and the first level dynamic redundancy register, wherein the pipeline bank is configured to: (a) write a data word into the memory bank at a selected one of the plurality of memory addresses; (b) verify the data word written into the memory bank to determine whether the data word was successfully written by the write; and (c) responsive to a determination that the data word was not successfully written by the write, writing the data word into the first level dynamic redundancy register, wherein the memory bank is fabricated on a first die and further wherein the first level dynamic redundancy register is fabricated on a second die.

Abstract: A method for fabricating an array of pillars. The method includes fabricating an MTJ (magnetic tunnel junction) film deposition metal stack on a CMOS wafer. The method selects between subsequent electron beam patterning for the wafer and photolithography patterning for the wafer. For electron beam patterning, an electron beam lithography hard mask is deposited onto the metal stack, and an electron beam is used to pattern a first array of pillars into the electron beam lithography hard mask to produce a first resulting pillar array. For photolithography patterning, a photolithography hard mask is deposited onto the metal stack, and photolithography is used to pattern a second array of pillars into the photolithography hard mask to produce a second resulting pillar array. The first resulting pillar array is substantially the same as the second resulting pillar array.

Abstract: A method for selectively writing to STT-MRAM using an AC current is provided. The method is performed in a memory device including two or more multilevel magnetic tunnel junctions (MTJs) arranged in series with respect to a single terminal of a transistor, where the two or more multilevel MTJs include a first MTJ having a first magnetic characteristic and first electrical characteristic and a second MTJ having a second magnetic characteristic that is distinct from the first magnetic characteristic and a second electrical characteristic. The method includes writing to an MTJ. The writing includes applying a DC current to the two or more MTJs and applying an AC current to the two or more MTJs, where the AC current is adjusted to a frequency that is tuned to a write assist frequency corresponding to the respective MTJ.