G11C5/05

CROSSBAR ARRAYS FOR CALCULATING MATRIX MULTIPLICATION
20180350433 · 2018-12-06 ·

A crossbar array, comprises a plurality of row lines, a plurality of column lines intersecting the plurality of row lines at a plurality of intersections, a plurality of junctions coupled between the plurality of row lines and the plurality of column lines at a portion of the plurality of intersections, and a plurality of diagonal control lines coupled to the plurality of junctions. Each junction comprises a resistive memory element and a transistor, and the junctions are positioned to calculate a matrix multiplication of a first matrix and a second matrix.

METHOD FOR DETERMINING A MEMORY WINDOW OF A RESISTIVE RANDOM ACCESS MEMORY
20180197603 · 2018-07-12 ·

A method for determining a memory window of at least one resistive random access memory cell, the resistive random access memory cell including a high resistance state and a low resistance state, the passage of the resistive random access memory from an initial state among the high resistance state or the low resistance state to another state then the return to the initial state forming a cycle, the method including: measuring the values of the resistances of the high resistance and low resistance states at a given cycle j, j being an integer; determining the memory window to use during the n cycles following the given cycle j, n being an integer, the memory window being calculated by taking into account at least the resistances of the high resistance and low resistance states at the cycle j.

DIMENSION SHUFFLING USING MATRIX PROCESSORS

In one embodiment, a matrix operation may be performed to reorder a plurality of dimensions of an input matrix stored in two-dimensional memory. Data associated with the input matrix may be accessed using one or more strided memory operations, wherein the one or more strided memory operations are configured to access the two-dimensional memory at a plurality of locations that are separated by a particular interval. The data accessed using the one or more strided memory operations may be stored in a result matrix, wherein the data accessed using each strided memory operation is stored in the result matrix in non-transpose form or transpose form.

Memory array with power-efficient read architecture
09842652 · 2017-12-12 · ·

Various embodiments comprise apparatuses and methods including a three-dimensional memory apparatus having upper strings and lower strings. The upper strings can include a first string of memory cells and a second string of memory cells arranged substantially parallel and adjacent to one another. The lower strings can include a third string of memory cells and a fourth string of memory cells arranged substantially parallel and adjacent to one another. The strings can each have a separate sense amplifier coupled thereto. The first and third strings and the second and fourth strings can be configured to be respectively coupled in series with each other during a read operation. Additional apparatuses and methods are described.

NAND flash memory device with oblique architecture and memory cell array

A memory device including multiple word lines, multiple bit lines and a memory cell array is provided. The word lines intersect the bit lines, and an included angle between the word lines and the bit lines is not a right angle. The memory cell array includes multiple memory cells respectively disposed at the intersections of the word lines and the bit lines. Each row of the memory cells is electrically connected to one of the word lines, and each column of the memory cells is electrically connected to one of the bit lines.