A logic implementation method based on bipolar memristor

Abstract

The invention discloses a logic implementation method based on a bipolar memristor. One input is defined as the initial resistance state of the bipolar memristor, the other input is defined as voltage, and the output is defined as the final resistance state of the device. Use one device to complete 16 kinds of Boolean logic functions within one-step read and two-step write operations, including XOR logic that is difficult to implement for other logic operations. It only needs one device to read and write once. This method does not require initialization and can directly perform logic cascading, which is more efficient than existing logic methods. Simultaneously, the present invention also provides a kind of realization scheme of one-bit full adder and two-bit multiplier based on this method, compared with other methods, the device number that this method uses is less, and number of steps is also less, so more efficient. Therefore, this method can be used as a general logic method with great development prospects.

Description

technical field [0001] The invention belongs to the field of new computing technologies, and in particular relates to a logic implementation method based on a bipolar memristor. Background technique [0002] Integrated circuits have developed rapidly in accordance with Moore's Law for half a century, but with the reduction in device size and gradually approaching the physical limit, Moore's Law is difficult to maintain. Simply shrinking the size of CMOS devices is difficult to continue to greatly improve performance, therefore, we need new materials, devices and architectures to meet the needs of the era of data explosion. Today's computers are based on the Von Neumann architecture. The separation of storage and computing in the Von Neumann architecture severely limits the performance of computers. This is the so-called Von Neumann bottleneck, a new architecture that combines storage and computing. It is expected to break through this bottleneck. [0003] In recent years, ...

AI Technical Summary

Problems solved by technology

Existing memristor-based logic implementation methods are mainly divided into two categories, one is using a voltage signal as two inputs and a resistance value as an output. It is impossible to directly perform logic cascading, and the output needs to be converted into a voltage signal through AD, and then used as an input again, and usually needs to be initialized, which greatly increases the number of steps required to implement logic operations; Value form, although this method is beneficial to logic cascading, but because the input and output are expressed in the form of resistance, the number of devices required is large

Therefore, there is currently no logic implementation method that can use a small number of devices, realize all logic operations in a few steps, and can be directly cascaded.

Images