Abstract: A method includes providing a substrate having a surface such that a first hard mask layer is formed over the surface and a second hard mask layer is formed over the first hard mask layer, forming a first pattern in the second hard mask layer, where the first pattern includes a first mandrel oriented lengthwise in a first direction and a second mandrel oriented lengthwise in a second direction different from the first direction, and where the first mandrel has a top surface, a first sidewall, and a second sidewall opposite to the first sidewall, and depositing a material towards the first mandrel and the second mandrel such that a layer of the material is formed on the top surface and the first sidewall but not the second sidewall of the first mandrel.

Abstract: A transpose memory unit for a plurality of multi-bit convolutional neural network based computing-in-memory applications includes a memory cell and a transpose cell. The memory cell stores a weight. The transpose cell is connected to the memory cell and receives the weight from the memory cell. The transpose cell includes an input bit line, at least one first input word line, a first output bit line, at least one second input word line and a second output bit line. One of the at least one first input word line and the at least one second input word line transmits at least one multi-bit input value, and the transpose cell is controlled by the second word line to generate a multiply-accumulate output value on one of the first output bit line and the second output bit line according to the at least one multi-bit input value multiplied by the weight.

Abstract: A transpose memory unit for a plurality of multi-bit convolutional neural network based computing-in-memory applications includes a memory cell and a transpose cell. The memory cell stores a weight. The transpose cell is connected to the memory cell and receives the weight from the memory cell. The transpose cell includes an input bit line, at least one first input word line, a first output bit line, at least one second input word line and a second output bit line. One of the at least one first input word line and the at least one second input word line transmits at least one multi-bit input value, and the transpose cell is controlled by the second word line to generate a multiply-accumulate output value on one of the first output bit line and the second output bit line according to the at least one multi-bit input value multiplied by the weight.