Abstract: Precharging circuits and techniques are presented for use with magnetic memory devices in order to speed up access to the memory cells for reading and writing. Including precharging in the sense amplifiers used to access the memory cells enables self-referenced read operations to be completed more quickly than is possible without precharging. Similarly, precharging can also be used in conjunction with write-back operations in order to allow the data state stored by magnetic tunnel junctions included in the memory cells to be changed more rapidly.

Abstract: A magnetoresistive-based device and method of manufacturing a magnetoresistive-based device using one or more hard masks. The process of manufacture, in one embodiment, includes patterning a mask, after patterning the mask, etching (a) through a first layer of electrically conductive material to form an electrically conductive electrode and (b) through a third layer of ferromagnetic material to provide sidewalls of the second synthetic antiferromagnetic structure. The process further includes providing insulating material on or over the sidewalls of the second synthetic antiferromagnetic structure and, thereafter, etching through (a) a second tunnel barrier layer to provide sidewalls thereof, (b) a second layer of ferromagnetic material to provide sidewalls thereof, (c) a first tunnel barrier layer to provide sidewalls thereof, and (d) a first layer of ferromagnetic material to provide sidewalls of the first synthetic antiferromagnetic structure.

Abstract: A magnetoresistive stack/structure and method of manufacturing same comprising wherein the stack/structure includes a seed region, a fixed magnetic region disposed on and in contact with the seed region, a dielectric layer(s) disposed on the fixed magnetic region and a free magnetic region disposed on the dielectric layer(s). In one embodiment, the seed region comprises an alloy including nickel and chromium having (i) a thickness greater than or equal to 40 Angstroms (+/?10%) and less than or equal to 60 Angstroms (+/?10%), and (ii) a material composition or content of chromium within a range of 25-60 atomic percent (+/?10%) or 30-50 atomic percent (+/?10%).

Abstract: In some examples, a memory device is configured with a reduced command set and a variable burst length. In some instances, the variable burst length defines a page size associated with data to be loaded into a cache. In other instances, the variable burst length may be set on the fly per read/write command and, in some cases, the burst length may be utilized to define the page size associated with the read/write command.

Abstract: Higher word line voltages facilitate write operations in spin-torque magnetic memory devices, but overdriving the gate of a selection transistor with such higher word line voltages can damage the selection transistor if the gate-to-source voltage for the selection transistor is too high. Therefore in order to support the word line voltage needed on the gate of the select transistor for an up-current write operation without exceeding limits on the gate-to-source voltage for the select transistor, the gate of the selection transistor can be driven in a two-step process. The gate of the selection transistor is first driven to a lower voltage within the limits of the gate-to-source voltage for the transistor when the source of the transistor is grounded or at a voltage near ground. A voltage is then applied across the memory cell, which results in the source of the selection transistor being raised above its initial ground or near-ground state.

Abstract: A memory having a delayed write-back to the array of data corresponding to a previously opened page allows delays associated with write-back operations to be avoided. After an initial activation opens a first page and the read/write operations for that page are complete, write-back of the open page to the array of memory cells is delayed until after completion of a subsequent activate operation that opens a new page. Techniques to force a write-back in the absence of another activate operation are also disclosed. Calibration and testing sequences are also supported in which a non-destructive mode preserves data stored in a non-volatile memory array and status bits used to indicate open pages are cleared so later inadvertent delayed write-back operations as a result of the calibration or testing do not corrupt the non-volatile data.

Abstract: In some examples, a memory is configured to write multiple pages of an internal page size from a cache on the memory to a memory array on the memory in response to receiving a single precharge command when in a page emulation mode. When in the page emulation mode, the memory also reads multiple pages of the internal page size from the memory array and stores them in the cache in response to receiving a single activate command.

Abstract: The present disclosure is drawn to, among other things, a method of fabricating an integrated circuit device having a magnetoresistive device. In some aspects, the method includes forming the magnetoresistive device on a first contact of a substrate, wherein the magnetoresistive device includes a fixed magnetic region and a free magnetic region separated by an intermediate region; depositing a first dielectric material over the magnetoresistive device; depositing a second dielectric material over the first dielectric material; polishing a surface of the second dielectric material; forming a first cavity through the polished surface of the second dielectric material to expose a surface of the magnetoresistive device; and depositing an electrically conductive material in the first cavity to form a via.

Abstract: In some examples, a nonvolatile storage element may be configured to store a state or value during a low power or powered down period of a circuit. For example, the nonvolatile storage element may include a bridge of resistive elements that have a resistive state that may be configured by applying voltages to multiple drive paths. A sense amplifier may be connected to the bridge in order to resolve a voltage differential associated with the bridge to ether power or ground and, thereby determine the state associated with on the nonvolatile storage element.

Abstract: Magnetoresistive device architectures and methods for manufacturing are presented that facilitate integration of process steps associated with forming such devices into standard process flows used for surrounding logic/circuitry. In some embodiments, the magnetoresistive device structures are designed such that the devices are able to fit within the vertical dimensions of the integrated circuit associated with a single metal layer and a single layer of interlayer dielectric material. Integrating the processing for the magnetoresistive devices can include using the same standard interlayer dielectric material as used in the surrounding circuits on the integrated circuit as well as using standard vias to interconnect to at least one of the electrodes of the magnetoresistive devices.

Abstract: A memory device includes one or more memory arrays and a quad bit error correction circuit. The quad bit error correction circuit may include a first layer error correction circuit and a second layer error correction circuit. The first layer error correction circuit may be configured to generate a Hamming correction bit vector, and the second layer error correction circuit may be configured to generate a Golay correction bit vector. The Hamming correction bit vector and the Golay correction bit vector may be used to identify up to four correctable bit errors in data to be stored in the one more memory arrays.

Abstract: A memory having a delayed write-back to the array of data corresponding to a previously opened page allows delays associated with write-back operations to be avoided. After an initial activation opens a first page and the read/write operations for that page are complete, write-back of the open page to the array of memory cells is delayed until after completion of a subsequent activate operation that opens a new page. Techniques to force a write-back in the absence of another activate operation are also disclosed.

Abstract: A method of manufacturing a magnetoresistive stack/structure comprising (a) etching through a second magnetic region to (i) provide sidewalls of the second magnetic region and (ii) expose a surface of a dielectric layer, (b) depositing a first encapsulation layer on the sidewalls of the second magnetic region and over a surface of the dielectric layer, (c) thereafter: (i) etching the first encapsulation layer which is disposed over the dielectric layer using a first etch process, and (ii) etching re-deposited material using a second etch process, wherein, after such etching, a portion of the first encapsulation layer remains on the sidewalls of the second magnetic region, (d) etching (i) through the dielectric layer to form a tunnel barrier and provide sidewalls thereof and (ii) etching the first magnetic region to provide sidewalls thereof, and (e) depositing a second encapsulation layer on the sidewalls of the tunnel barrier and first magnetic region.

Abstract: The present disclosure is drawn to, among other things, a method of managing a memory device. In some aspects, the method includes scanning a first memory region for bit errors; in response to detecting one or more bit errors in the first memory region, incrementing a counter associated with the first memory region based on the number of bit errors detected; comparing a total number of bit errors against a threshold, wherein the total number of bit errors is identified from the first counter; and, if the total number of bit errors exceeds the threshold, restricting access to the first memory region by mapping an address corresponding to the first memory region to a second memory region.

Abstract: A magnetoresistive-based device and method of manufacturing a magnetoresistive-based device using one or more hard masks. The process of manufacture, in one embodiment, includes patterning a mask, after patterning the mask, etching (a) through a first layer of electrically conductive material to form an electrically conductive electrode and (b) through a third layer of ferromagnetic material to provide sidewalls of the second synthetic antiferromagnetic structure. The process further includes providing insulating material on or over the sidewalls of the second synthetic antiferromagnetic structure and, thereafter, etching through (a) a second tunnel barrier layer to provide sidewalls thereof, (b) a second layer of ferromagnetic material to provide sidewalls thereof, (c) a first tunnel barrier layer to provide sidewalls thereof, and (d) a first layer of ferromagnetic material to provide sidewalls of the first synthetic antiferromagnetic structure.

Abstract: Once a delay locked loop has been locked to a clock signal, an omitted clock cycle is injected into the input of the delay locked loop without stopping the operation of the delay locked loop. The omitted cycle is later detected at an output of the delay locked loop, and the delay between the input and output is determined based on the time the omitted cycle requires to propagate through the delay locked loop. Once determined, the number of cycles of delay for the delay locked loop can be used in conjunction with an internal clock signal to launch data and/or data strobes from memory devices and memory controllers such that the proper phase alignment and clock cycle alignment is achieved.

Abstract: Methods for manufacturing magnetoresistive devices are presented in which isolation of magnetic layers in the magnetoresistive stack is achieved by oxidizing exposed sidewalls of the magnetic layers and then depositing additional encapsulating material prior to subsequent etching steps. Etching the magnetic layers using a non-reactive gas further prevents degradation of the sidewalls.

Abstract: A magnetoresistive element (e.g., a spin-torque magnetoresistive memory element) includes a fixed magnetic layer, a free magnetic layer, having a high-iron alloy interface region located along a surface of the free magnetic layer, wherein the high-iron alloy interface region has at least 50% iron by atomic composition, and a first dielectric, disposed between the fixed magnetic layer and the free magnetic layer. The magnetoresistive element further includes a second dielectric, having a first surface that is in contact with the surface of the free magnetic layer, and an electrode, disposed between the second dielectric and a conductor. The electrode includes: (i) a non-ferromagnetic portion having a surface that is in contact with a second surface of the second dielectric, and (ii) a second portion having at least one ferromagnetic material disposed between the non-ferromagnetic portion of the electrode and the conductor.

Abstract: Techniques for configuring the layers included in the free portion of a spin-torque magnetoresistive device are presented that allow for characteristics of the free portion to be tuned to meet the needs of various applications. In one embodiment, high data retention is achieved by balancing the perpendicular magnetic anisotropy of the ferromagnetic layers in the free portion. In other embodiments, imbalanced ferromagnetic layers provide for lower switching current for the magnetoresistive device. In various embodiments, different coupling layers can be used to provide exchange coupling between the ferromagnetic layers in the free portion, including oscillatory coupling layers, ferromagnetic coupling layers using materials that can alloy with the neighboring ferromagnetic layers, and discontinuous layers of dielectric material such as MgO that result in limited coupling between the ferromagnetic layers and increases perpendicular magnetic anisotropy (PMA) at the interface with those layers.

Abstract: In a spin-torque magnetic random access memory (MRAM) that includes local source lines, auto-booting of the word line is used to reduce power consumption by reusing charge already present from driving a plurality of bit lines during writing operations. Auto-booting is accomplished by first driving the word line to a first word line voltage. After such driving, the word line isolated. Subsequent driving of the plurality of bit lines that are capacitively coupled to the word line causes the word line voltage to be increased to a level desired to allow sufficient current to flow through a selected memory cell to write information into the selected memory cell. Additional embodiments include the use of a supplemental voltage provider that is able to further boost or hold the isolated word line at the needed voltage level.