An in-memory computing architecture is disclosed that can evaluate the transitive closure of graphs using the natural parallel flow of information in 3-D nanoscale crossbars. The architecture can be implemented using 3-D crossbar architectures with as few as two layers of 1-diode 1-resistor (1D1R) interconnects. The architecture avoids memory-processor bottlenecks and can hence scale to large graphs. The approach leads to a runtime complexity of O(n.sup.2) using O(n.sup.2) memristor devices. This compares favorably to conventional algorithms with a time complexity of O((n.sup.3)/p+(n.sup.2) log p) on p processors. The approach takes advantage of the dynamics of 3-D crossbars not available on 2-D crossbars.

Methods, apparatuses, and systems for in-or near-memory processing are described. Bits of a first number may be stored on a number of memory elements, wherein each memory element of the number of memory elements intersects a bit line and a word line of a number of word lines. A number of signals corresponding to bits of a second number may be driven on the number of word lines to generate a number of output signals. A value equal to a product of the first number and the second number may be generated based on the number of output signals.

An information processing device, including a resistive analog neuromorphic device element having a pair of electrodes and an oxide layer provided between the pair of electrodes, and a parallel circuit having a low resistance component and a capacitance component. The parallel circuit and the resistive analog neuromorphic device element are connected in series.

The invention provides a microelectronic device comprising at least two memory cells each comprising a so-called selection transistor and a memory element associated with said selection transistor, each transistor comprising a channel in the form of a wire extending in a first direction (x), a gate bordering said channel, a source extending in a second direction (y), and a drain connected to the memory element, said transistors being stacked in a third direction (z) and each occupying a given altitude level in the third direction (z), the microelectronic device wherein the source and the drain are entirely covered by spacers projecting in the third direction (z) in a plane (xy). The invention also provides a method for manufacturing such a device.

Technologies relating to increasing the surface area of selectors in crossbar array circuits are provided. An example apparatus includes: a substrate; a first line electrode formed on the substrate; an RRAM stack formed on the first line electrode, wherein the RRAM stack; an isolation layer formed beside the RRAM stack, wherein the isolation layer includes an upper surface and a sidewall, and a height from the upper surface to the first line electrode is 100 nanometers to 10 micrometers; a selector stack formed on the RRAM stack, the sidewall, and the upper surface; and a second line electrode formed on the selector stack.

Memories, memory controllers, and computing systems and their methods of operation are disclosed. In some embodiments, a method of accessing a memory includes accessing a first bit line corresponding to a sense amplifier and accessing a second bit line corresponding to the sense amplifier. In some embodiments, a memory controller includes a second memory configured to store data of a second data type. In some embodiments, a method includes operating a memory in a second mode in response to receiving an input to change the operation of the memory from a first mode to the second mode.

A DIMA semiconductor structure is disclosed. The DIMA semiconductor structure includes a frontend including a semiconductor substrate, a transistor switch of a memory cell coupled to the semiconductor substrate and a computation circuit on the periphery of the frontend coupled to the semiconductor substrate. Additionally, the DIMA includes a backend that includes an RRAM component of the memory cell that is coupled to the transistor switch.

Various embodiments of the present disclosure are directed towards a memory cell comprising a high electron affinity dielectric layer at a bottom electrode. The high electron affinity dielectric layer is one of multiple different dielectric layers vertically stacked between the bottom electrode and a top electrode overlying the bottom electrode. Further, the high electrode electron affinity dielectric layer has a highest electron affinity amongst the multiple different dielectric layers and is closest to the bottom electrode. The different dielectric layers are different in terms of material systems and/or material compositions. It has been appreciated that by arranging the high electron affinity dielectric layer closest to the bottom electrode, the likelihood of the memory cell becoming stuck during cycling is reduced at least when the memory cell is RRAM. Hence, the likelihood of a hard reset/failure bit is reduced.

This method for training a binarized neural network, also called BNN, including neurons, with a binary weight for each connection between two neurons, is implemented by an electronic circuit and comprises: a forward pass including calculating an output vector by applying the BNN on an input vector; a backward pass including computing an error vector from the calculated output vector, and calculating a new value of the input vector by applying the BNN on the error vector; a weight update including computing a product by multiplying an element of the error vector with one of the new value of the input vector, modifying a latent variable depending on the product; and updating the weight with the latent variable;

each weight being encoded using a primary memory component;

each latent variable being encoded using a secondary memory component having a characteristic subject to a time drift.

The present invention is directed to a magnetic memory cell including a magnetic tunnel junction (MTJ) memory element and a two-terminal bidirectional selector coupled in series between two conductive lines. The MTJ memory element includes a magnetic free layer; a magnetic reference layer; and an insulating tunnel junction layer interposed therebetween. The two-terminal bidirectional selector includes a bottom electrode; a top electrode; a load-resistance layer interposed between the bottom and top electrodes and comprising a first tantalum oxide; a first volatile switching layer interposed between the bottom and top electrodes and comprising a metal dopant and a second tantalum oxide that has a higher oxygen content than the first tantalum oxide; and a second volatile switching layer in contact with the first volatile switching layer and comprising a third tantalum oxide that has a higher oxygen content than the first tantalum oxide.