Memory devices provide a plurality of memory cells, each memory cell including a memory element and a selection device. A plurality of first (e.g., row) address lines can be adjacent (e.g., under) a first side of at least some cells of the plurality. A plurality of second (e.g., column) address lines extend across the plurality of row address lines, each column address line being adjacent (e.g., over) a second, opposing side of at least some of the cells. Control circuitry can be configured to selectively apply a read voltage or a write voltage substantially simultaneously to the address lines. Systems including such memory devices and methods of accessing a plurality of cells at least substantially simultaneously are also provided.

1. The present disclosure is drawn to, among other things, a method for accessing memory using dual standby modes, the method including receiving a first standby mode indication selecting a first standby mode from a first standby mode or a second standby mode, configuring a read bias system to provide a read bias voltage and a write bias system to provide approximately no voltage, or any voltage outside the necessary range for write operation, based on the first standby mode, receiving a second standby mode indication selecting the second standby mode, and configuring the read bias system to provide at least the read bias voltage and the write bias system to provide a write bias voltage based on the second standby mode, the read bias voltage being lower than the write bias voltage.

Concurrent access of multiple memory cells in a cross-point memory array is disclosed. In one aspect, a forced current approach is used in which, while a select voltage is applied to a selected bit line, an access current is driven separately through each selected word line to concurrently drive the access current separately through each selected memory cell. Hence, multiple memory cells are concurrently accessed. In some aspects, the memory cells are accessed using a self-referenced read (SRR), which improves read margin. Concurrently accessing more than one memory cell in a cross-point memory array improves bandwidth. Moreover, such concurrent accessing allows the memory system to be constructed with fewer, but larger cross-point arrays, which increases array efficiency. Moreover, concurrent access as disclosed herein is compatible with memory cells such as MRAM which require bipolar operation.

A non-volatile memory device includes arrays of non-volatile memory cells that are configured to the store weights for a recurrent neural network (RNN) inference engine with a gated recurrent unit (GRU) cell. A set three non-volatile memory arrays, such as formed of storage class memory, store a corresponding three sets of weights and are used to perform compute-in-memory inferencing. The hidden state of a previous iteration and an external input are applied to the weights of the first and the of second of the arrays, with the output of the first array used to generate an input to the third array, which also receives the external input. The hidden state of the current generation is generated from the outputs of the second and third arrays.

A method of testing a non-volatile memory (NVM) array includes obtaining a current distribution of a subset of NVM cells of the NVM array, the current distribution including first and second portions corresponding to respective logically high and low states of the subset of NVM cells, programming an entirety of the NVM cells of the NVM array to one of the logically high or low states, determining an initial bit error rate (BER) by performing first and second pass/fail (P/F) tests on each NVM cell of the NVM array, and using the current distribution to adjust the initial BER rate. Each of obtaining the current distribution, programming the entirety of the NVM cells, and performing the first and second P/F tests is performed while the NVM array is heated to a target temperature.

A semiconductor device including a magnetic tunnel junction (MTJ) stack, a first metal line above the MTJ stack and a magnetoelectric material layer above the first metal line. A semiconductor device including an array of magnetic tunnel junction (MTJ) stacks, a first metal line connected physically and electrically to a top electrode of each MTJ stack in a row of the array of MTJ stacks and a magnetoelectric material layer above the first metal line, connected physically and electrically to the first metal line. A method including forming an array of magnetic tunnel junction (MTJ) stacks, forming a first metal line above a row of the array of MTJ stacks, and forming a magnetoelectric material layer above the first metal line, connected physically and electrically to the first metal line.

A magnetic multilayer film for a magnetic memory element includes an amorphous heavy metal layer having a multilayer structure in which a plurality of first layers containing Hf alternate repeatedly with a plurality of second layers containing a heavy metal excluding Hf; and a recording layer that includes a ferromagnetic layer and that is adjacent to the heavy metal layer, the ferromagnetic layer having a variable magnetization direction.

The present disclosure is drawn to, among other things, a configuration bit including at least four resistive elements and a voltage amplifier. At least two first resistive elements may be electrically connected in series via a first electrode and at least two second resistive elements may be electrically connected in series via a second electrode. The at least two first resistive elements may be electrically connected in parallel to the at least two second resistive elements via a third electrode and a fourth electrode. The first electrode and the second electrode may be electrically connected to a voltage supply. The third electrode and the fourth electrode may be electrically connected to an input of the voltage amplifier.

A semiconductor device capable of changing a data programming process in a simple manner according to a situation is provided. The semiconductor device includes a plurality of memory cells, a programming circuit for supplying a programming current to the memory cell, and a power supply circuit for supplying power to the programming circuit. The power supply circuit includes a charge pump circuit for boosting the external power supply, a voltage of the external power supply according to the selection indication, and a selectable circuit capable of switching the boosted voltage boosted by the charge pump circuit. The control circuit further includes a control circuit for executing data programming processing by the programming circuit by switching the selection indication.

A memory device includes a spin-orbit-transfer (SOT) bottom electrode, an SOT ferromagnetic free layer, a first tunnel barrier layer, a spin-transfer-torque (STT) ferromagnetic free layer, a second tunnel barrier layer and a reference layer. The SOT ferromagnetic free layer is over the SOT bottom electrode. The SOT ferromagnetic free layer has a magnetic orientation switchable by the SOT bottom electrode using a spin Hall effect or Rashba effect. The first tunnel barrier layer is over the SOT ferromagnetic free layer. The STT ferromagnetic free layer is over the first tunnel barrier layer and has a magnetic orientation switchable using an STT effect. The second tunnel barrier layer is over the STT ferromagnetic free layer. The second tunnel barrier layer has a thickness different from a thickness of the first tunnel barrier layer. The reference layer is over the second tunnel barrier layer and has a fixed magnetic orientation.